Molecular Plant Advance Access originally published online on August 10, 2009
Molecular Plant 2009 2(5):977-989; doi:10.1093/mp/ssp059
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Plant Cell Wall Proteomics: Mass Spectrometry Data, a Trove for Research on Protein Structure/Function Relationships
Surfaces Cellulaires et Signalisation chez les Végétaux, UMR 5546 CNRS–UPS–Université de Toulouse, Pôle de Biotechnologie Végétale, 24 chemin de Borde-Rouge, BP 42617 Auzeville, 31326 Castanet-Tolosan, France
1 To whom correspondence should be addressed. E-mail jamet{at}scsv.ups-tlse.fr, fax +33 562 193 502, tel. +33 562 193 530.
 |
Abstract |
|---|
 TOP Abstract INTRODUCTIONRESULTS AND DISCUSSIONMethodsSUPPLEMENTARY DATAFUNDING |
|---|
Proteomics allows the large-scale study of protein expression either in whole organisms or in purified organelles. In particular, mass spectrometry (MS) analysis of gel-separated proteins produces data not only for protein identification, but for protein structure, location, and processing as well. An in-depth analysis was performed on MS data from etiolated hypocotyl cell wall proteomics of Arabidopsis thaliana. These analyses show that highly homologous members of multigene families can be differentiated. Two lectins presenting 93% amino acid identity were identified using peptide mass fingerprinting. Although the identification of structural proteins such as extensins or hydroxyproline/proline-rich proteins (H/PRPs) is arduous, different types of MS spectra were exploited to identify and characterize an H/PRP. Maturation events in a couple of cell wall proteins (CWPs) were analyzed using site mapping. N-glycosylation of CWPs as well as the hydroxylation or oxidation of amino acids were also explored, adding information to improve our understanding of CWP structure/function relationships. A bioinformatic tool was developed to locate by means of MS the N-terminus of mature secreted proteins and N-glycosylation.
Key Words: Cell wall protein • MALDI–TOF • mass spectrometry • post-translational modification • protein structure • proteomics
Received for publication May 4, 2009. Accepted for publication July 13, 2009.
|
INTRODUCTION |
|---|
|
TOPAbstractINTRODUCTIONRESULTS AND DISCUSSIONMethodsSUPPLEMENTARY DATAFUNDING |
|---|
The plant cell wall is an essential cell compartment that has structural as well as functional roles. The envelope of the plant cell is not only responsible for its shape; it is also a dynamic structure in direct contact with the environment. Mostly composed of polysaccharides (up to 95% in mass), the cell wall also contains proteins and, in some cell types, lignins (Carpita and Gibeaut, 1993). The composition and organization of polysaccharides have been widely described. The global architecture of this network is now well studied, and several models are available (Willats et al., 2001; Fry, 2004; Somerville et al., 2004). Among cell wall proteins (CWPs), structural proteins such as extensins and hydroxyproline/proline-rich proteins (H/PRPs) are known to reinforce the polysaccharide networks (Showalter, 1993; Kieliszewski and Lamport, 1994). From a functional point of view, the cell wall is essential for growth and differentiation, signaling, and response to biotic and abiotic stresses (Roberts, 1990, 2001; Ellis et al., 2002; Vogel et al., 2004). Since the true actors of cell wall dynamics are proteins, all CWPs other than structural proteins are of interest. Therefore, to better understand the cell wall complexity, the challenge is to go further into the identification of the CWPs and their structure/function relationships. In this context, the last few years saw the emergence of cell wall proteomics aimed at profiling CWPs at a given time in specific environmental conditions. Beyond some studies performed on Medicago sativa (Watson et al., 2004), Zea mays (Alvarez et al., 2006; Zhu et al., 2006), Cicer arietinum (Bhushan et al., 2006), and Oryza sativa (Chen et al., 2008; Jung et al., 2008; Cho et al., 2009), most data on plant cell wall proteomics concern the model plant Arabidopsis thaliana. The knowledge on CWPs has recently been reviewed and updated (Jamet et al., 2006, 2008). As an active contributor to the description of A. thaliana CWPs (Borderies et al., 2003; Boudart et al., 2005; Charmont et al., 2005; Feiz et al., 2006; Minic et al., 2007; Irshad et al., 2008), our group has built a database (WallProtDB) containing all A. thaliana and O. sativa CWPs identified in different organs. So far, this database contains nearly 800 CWPs and is freely accessible to the public (www.polebio.scsv.ups-tlse.fr/WallProtDB/). CWPs in the database were classified into nine functional classes according to bioinformatic predictions, thus revealing a great diversity of protein functions inside the cell wall. We now have a good picture of CWPs present in the plant cell wall, but the exact functions of most of these proteins are still unknown. Several descriptions of CWP mutants are proof of the importance of proteins in cell walls both during plant development and in response to pathogens (Berger and Altmann, 2000; Baumberger et al., 2003; Vogel et al., 2004; Roudier et al., 2005; Kurdyukov et al., 2006). However, many topics related to CWPs still need to be addressed to achieve a better understanding of cell wall dynamics. Among these topics are: the role of different members of multigenic families, the identification of substrates or partners, the post-translational modifications (PTMs), and the spatial and temporal regulation.
Several strategies can be used in cell wall proteomic studies (Jamet et al., 2008). A simplified flowchart for plant cell wall proteomics is presented in Figure 1. Defining a specific strategy to answer a biological question requires two essential decisions involving: (1) the choice of the appropriate material; and (2) a method to prepare a protein extract containing as few intracellular proteins as possible (Feiz et al., 2006). Since CWPs are mostly basic glycoproteins, a good protein separation was obtained simply by mono-dimensional electrophoresis (1D-E) (Boudart et al., 2005). However, to improve CWP fractionation, cationic exchange chromatography (CEC) performed in acidic pH was shown to be a method of choice, because it mimics the physico-chemical conditions in cell walls (Irshad et al., 2008). In a second step, chromatography fractions were separated by 1D-E, thus offering a more suitable bi-dimensional (2D) separation than 2D-E. The staining of gels can be performed in different ways, revealing different types of proteins with different sensitivities (Chevalier et al., 2004; Jamet et al., 2008). Stained bands of interest were excised from gels and subsequently submitted to protease digestion prior to analysis by MS. Two approaches are widely used for protein identification: (1) peptide mass fingerprinting (PMF) (MS spectrum); and (2) peptide sequencing by tandem mass spectrometry (MS/MS spectrum) (Aebersold and Mann, 2003). The main advantage of the flowchart in Figure 1 is that it provides samples containing whole proteins ready for MS analysis. Analysis of complex samples was also performed, but is more difficult to interpret because the peptides of all the proteins are mixed (Bayer et al., 2006). Reconstruction of whole proteins requires more powerful software, and the structural information for a given protein is more difficult to obtain. Each step described in Figure 1 can be optimized to yield better results. In addition, MS technologies have evolved very quickly these last years, and they offer a large choice of sources of ionization and of analyzers (Aebersold and Mann, 2003; Han et al., 2008). However, to identify and characterize proteins, MALDI–TOF MS and MALDI–TOF–TOF MS/MS remain powerful, easy-to-use methods. They are the only two techniques to be considered in this paper. Finally, the bioinformatic analysis is critical because it allows not only protein identification, but also prediction of protein sub-cellular location and functional domains (San Clemente et al., 2009).
|
Although proteomics has been shown to be a very valuable way to identify CWPs, mass spectrometry (MS) data can yield more information on protein structure, location of PTMs, and processing, thus tackling biochemical protein complexity and revealing the hidden part of the iceberg. This article seeks to reveal the potential capacity of MS to improve our knowledge of CWPs. Relevant examples were chosen from a previous cell wall proteomic study in etiolated A. thaliana hypocotyls, where 137 CWPs were identified (Irshad et al., 2008). A detailed analysis of the data allowed the discrimination of close members of multigenic families, highlighting of maturation events, locating of PTMs, or access of proteins that are recalcitrant under standard conditions.
|
RESULTS AND DISCUSSION |
|---|
|
TOPAbstractINTRODUCTIONRESULTS AND DISCUSSIONMethodsSUPPLEMENTARY DATAFUNDING |
|---|
Discrimination among Close Members of Multigenic Families
Most plant genes, including those encoding CWPs, belong to multigene families. These can be very large, and they can include closely related members. For example, in A. thaliana, there are 73 predicted peroxidase genes (Oliva et al., 2009), 66 pectin methylesterase (PME) genes (http://cellwall.genomics.purdue.edu/families/4-5-1.html), 33 xyloglucan endotransglucosylase/hydrolase genes (Rose et al., 2002), and 66 polygalacturonase genes (http://cellwall.genomics.purdue.edu/families/4-3-3.html). It is usually difficult to study the regulation of close members of these gene families at the protein level. Classical approaches using immunology are not efficient because of high amino acid similarity, but PMF using MALDI–TOF MS provides a powerful alternative. Indeed, subtle mass differences between peptides can be measured, the accuracy of mass determination being as small as 20 ppm (parts per million). As an example, the identification in the same sample of two closely related lectin homologs is reported in Figure 2. At1g78850 and At1g78860 share 93% amino acid identity. On MS spectra, seven peptides common to both proteins were found (boxed peptides) whereas four and three peptides were, respectively, specific to At1g78850 (underlined in green) and At1g78860 (underlined in blue) (Supplemental Figure 1). Note that in one case, the two versions of a peptide could be easily distinguished (ILENGNMVIYDSS/NGK), namely a difference of 11.02 Da in mass (1655.79 and 1666.81 Da, respectively). In our study, all CWPs could be identified unambiguously, but some intracellular contaminants of our cell wall preparation could not be discriminated because they belong to multigene families comprising closely related members. Either there was not enough protein coverage to find specific peptides, or the amino acid sequences of the proteins were too close, such as malate dehydrogenase (88% shared identity between At1g53240 and At3g15020), RUBISCO small subunit (96 or 97% shared identity among At5g38410, At5g38420, and At5g38430), actin (100% shared identity between At2g37620 and At3g53750). The last case was that of proteins with repetitive domains, such as polyubiquitins and ubiquitin extension proteins (e.g. At3g52590, At2g36170, At2g35635, and At1g31340).
|
A Site Mapping Approach to Post-Translational Maturation Events
To become biologically active, some enzymes need to lose a pro-sequence, and are submitted to post-translational maturation events. PMF can be used as a tool to show such events by mapping the identified peptides on the sequence of the proteins. To be reliable, it must be carried out on several independent samples. Two examples of possible CWP maturation events are shown in Figure 3.
|
In our proteomic study on etiolated A. thaliana hypocotyls, six different putative pectin methylesterases (PMEs) were identified (Irshad et al., 2008). Five of them are type-I PMEs with both a predicted N-terminal pectinesterase inhibitor domain (PMEI, PF04043), and a C-terminal pectinesterase domain (PME, PF01095) (Micheli, 2001). The cleavage of the pro-sequence was shown to depend on the presence of one or two conserved basic tetrad motifs (Wolf et al., 2009). In none of the cases did we find peptides in the PMEI domains. In seven independent identifications of At1g53830 (Figure 3A), all the peptides were mapped to the predicted PME domain, namely between amino acids 275 and 548, thus allowing a coverage of 47% of this region (Supplemental Figure 2). In addition, the apparent molecular masses of these PMEs after 1D-E in denaturing conditions was around 35 kDa, consistent with a maturation event of the protein. For a long time, it was thought that there was no PMEI domain in cell wall PMEs. It was assumed that PMEI domains of PMEs could inhibit their enzyme activity, thus preventing the premature de-esterification of pectins before their secretion (Micheli, 2001). However, it was recently shown that maturation of PMEs inside the secretory pathway is required for their secretion (Bosch et al., 2005; Wolf et al., 2009). Our results are consistent with the single presence of mature PMEs in cell walls. They strongly support the assumption that maturation occurs before protein secretion. However, this must be confirmed by Edman degradation sequencing to provide the actual N-terminus of the processed protein.
The second example of protein maturation is provided by 
-xylosidase At1g68560. In our samples, -xylosidase was repeatedly identified as the only component present in three wide Coomassie blue stained bands obtained after 1D-E in denaturing conditions (H1, H2, H3 in Figure 3B) (Irshad et al., 2008). Apparent molecular masses were between 75 and 100 kDa. Cartography of identified peptides on 18 independent fractions indicates that proteins in H3 lack N-terminal peptides (Supplemental Figure 3). Differences in molecular masses between H1, H2, and H3 are thus probably due either to the N-terminal processing of the protein, or to N-glycosylations. In fact, there are eight predicted N-glycosylation sites in the protein. We found that at least one of these sites is occupied. Stars in Figure 3B point to N-glycosylated forms of the protein on the SNHETLFNTTSSLVFK peptide located between amino acids 156 and 173. In addition, At1g68560 was identified in a previous proteomic study performed on a stem protein fraction retained on Concanavalin A (Minic et al., 2007). We suspect that different states of N-glycosylation may occur on the other predicted sites. Furthermore, our MS data were compared to results of the N-terminal sequencing of a cabbage -xylosidase, which is also a member of Brassicaceae. This enzyme shows 75% identity with At1g68560 and an apparent molecular mass of 85 kDa after 1D-E in denaturing conditions (Sampedro et al., 2001). Edman sequencing showed that this N-terminal sequence was ISGSELTFSYTTDPFSFAVKRRL (J. Sampedro, personal communication). It shares 87% identity with an At1g68560 peptide located between amino acids 123 and 146. The removal of the peptide upstream of this sequence would cause a mass shift of 10.9 kDa, which fits in with the difference observed between molecular masses of H2 and H3. Altogether, it suggests that At1g68560 is processed at its N-terminus in cell walls. The processing of cell wall enzymes that act on polysaccharides was already shown. An -L-arabinofuranosidase (ARAI) and a β-D-xylosidase (XYL) from barley are post-translationally processed at their C-terminus (Lee et al., 2003). Such C-terminal processing was also suggested for the β-D-xylosidases AtXYL1 and AtXYL4 (Minic et al., 2004). In this case, not all the C-termini of the proteins were covered by MALDI–TOF MS PMF. An endo-β-1,4-xylanase (XYN-1) of barley aleurone was shown to be processed both at its N- and C-termini (Caspers et al., 2001). In addition, cell wall proteomic studies have shown that proteases represent about 11.6% of identified CWPs, none of which has known substrates (Jamet et al., 2008). In conclusion, our data support the hypotheses of an N-terminal processing of the At1g68560 -xylosidase in muro and of the presence of N-glycans.
In order to facilitate MS data analysis and easily identify post-translational maturation or modification events, we developed a bioinformatic tool called ProTerNyc, which is freely accessible (www.polebio.scsv.ups-tlse.fr/ProTerNyc/). This software predicts the location of the N-terminus in mature secreted protein and the N-glycosylated peptides using the data obtained by MS. The N-terminus is predicted by TargetP (Emanuelsson et al., 2007). With this tool, it is possible to overcome the limitations of searches in databases that do not take into account the signal peptide cut or peptide glycosylation. After running ProTerNyc on a given protein sequence and on a mass list generated by MS experiments, it can be seen that some observed peaks can be attributed to putative N-terminal peptide and N-glycopeptides.
Location of Post-Translational Modifications
Aside from all the PMF data used for the cartography of a protein sequence, it is sometimes very instructive to go further into the MS spectra. When exploring the MS data for etiolated hypocotyl CWP fractions (Irshad et al., 2008), we observed some significant peaks not always used for protein identification. A detailed analysis of these data revealed that they are consistent with PTMs such as N-glycosylation or amino acid hydroxylation or oxidation.
N-glycosylations
Several peroxidases were found (Irshad et al., 2008). Figure 4A shows the MS spectrum allowing the identification of one of these peroxidases, encoded by At3g32980. Among the major peaks displayed in this spectrum, five were found to fit with non-modified peptides of At3g32980, leading to its identification. Other peaks in this spectrum were shown to belong to other proteins also identified in this sample, except for four peaks of high m/z (2661.16, 2864.25, 3685.54, and 3888.65). Since At3g32980 carries four putative N-glycosylation sites, we looked for possible glycopeptides. Using the m/z of the four peaks listed above (red arrows in Figure 4A) and those of canonical N-glycans observed in plants (Faye et al., 2005), we calculated that each of these four peaks may correspond to two modified peptides, P1 and P2. Both peptides possess one N-glycosylation consensus sequence and are supposed to carry two different complex motifs of N-glycosylation. Interestingly, in their unmodified form, P1 and P2 were absent from the spectrum, suggesting that in our experimental conditions, these glycosylation sites are always occupied. To confirm that these four peaks correspond to glycopeptides, we analyzed a total protein extract after hydrogen fluoride (HF) deglycosylation. MS analysis of the corresponding sample showed that all four peaks were absent (Figure 4B). This suggests that glycosylations were removed during the HF treatment, confirming the nature of these modifications.
|
It has been reported that N-glycans are asymmetrically distributed on soybean seed coat peroxidase, whereas they are more uniformly distributed over horseradish peroxidase (Gray, 1998; Gray and Montgomery, 2006). In addition, a study performed on cationic peanut peroxidase showed that N-glycans are essential to attain full catalytic ability and thermostability, and to influence protein folding (Lige et al., 2001). Other examples demonstrate the importance of protein N-glycosylation in plants. The presence of N-glycans on the jack bean
-mannosidase was shown to influence the enzyme activity as well as its oligomerization (Kimura et al., 1999). It was also proposed that N-glycosylation motifs on a plant cell wall polygalacturonase inhibitor protein (PGIP) may favor interaction with its fungal polygalacturonase target during host–pathogen interaction (Lim et al., 2009). Finally, several studies on defective glycosylation mutants have shown that the N-glycosylation pathway regulates protein quality control, salt tolerance, cell wall deposition, development, and plant–pathogen interactions (Pattison and Amtmann, 2009). Our results show that MS can be used to predict glycosylation and to gain new insight into the structural and functional characterization of proteins. However, until now, it was difficult to recognize N-glycopeptides using a systematic proteomic approach. Software such as ProteinProspector (http://prospector.ucsf.edu/) and MASCOT (www.matrixscience.com/home.html) provide a series of simple PTM options to search for peptides with modified masses, including modified amino acids. In-depth PTM analysis, such as glycosylations, requires the use of specific tools, since this information is not yet referenced in databases. For a given protein, the ProTerNyc tool (www.polebio.scsv.ups-tlse.fr/ProTerNyc/) described above can be used to calculate all putative N-glycopeptides and to identify them in a mass list obtained from MS experiments. It should be noted that mass tolerance for peptides with high m/z may need to be increased, since accuracy of mass determination decreases as mass increases. This software will undoubtedly be a precious tool to search for glycopeptides.
Amino Acid Hydroxylation or Oxidation
Many CWPs contain hydroxylated Pro (Hyp), which can be O-glycosylated (Kieliszewski and Lamport, 1994). The At3g16850 polygalacturonase was identified by PMF using 12 peptides among CWPs from etiolated hypocotyls (Irshad et al., 2008). Among these peptides, one (P peptide) was predicted to exist under several forms. Indeed, the mass spectrum displays an isotopic massif with a monoisotopic peak at m/z 2522.23, associated with two isotopic massifs at a distance of 16 and 32 Da, respectively (Figure 5). Such modifications may respectively correspond to the oxidation or hydroxylation of one or two amino acids. The P peptide has a theoretical monoisotopic mass of 2522.27 Da (Figure 5). We performed MALDI–TOF–TOF MS/MS experiments on each of these peptides in order to confirm that these three isotopic massifs correspond to the same peptide, and locate oxidations or hydroxylations of amino acids. A clear fragmentation pattern was obtained from the P peptide, confirming the proposed sequence (Supplemental Figure 4). Both MS/MS spectra of the modified forms (m/z 2583.23 and 2554.21) present the two most abundant fragments previously observed for the P peptide (y10 and y15, Supplemental Figure 5). However, one of these fragments (y15) differed by 16 or 32 Da from the corresponding peaks found on the P peptide fragmentation pattern. These results confirmed the presence of oxidations or hydroxylations and allowed their localization in a region of five possible amino acid residues (PSHWD, in bold in Figure 5). The oxidized or hydroxylated amino acids in this sequence probably are the Pro, His, Trp, and Asp, which, respectively, give rise to Hyp, oxohistidine, hydroxytryptophan, and hydroxyaspartate. Hyp is widespread in the plant kingdom. Pro hydroxylation is a PTM performed by prolyl 4-hydroxylases in the endoplasmic reticulum (Lamport, 1965). A code based on primary sequences was proposed for Pro hydroxylation and consecutive O-glycosylation in Hyp-rich glycoproteins (HRGPs) and in Hyp/Pro-rich proteins (H/PRPs) (Shpak et al., 1999; Kieliszewski, 2001). In the case of At3g16850, only one Pro may be hydroxylated in the P peptide within the Asp–Pro–Ser context, which has never been described in plant proteins. In addition, the presence of Hyp in a CWP different from the well known HRGPs and H/PRPs is noteworthy. A few cases have been reported in the literature, such as that of the dodeca-CLE peptides involved in cell differentiation (Ito et al., 2006; Kondo et al., 2006). Hyp was found in the Val–Pro–Ser and Gly–Pro–Asn sequences. So far, Pro hydroxylation in the CLE peptides seems to have no effect on their biological activity. Unlike Hyp, the presence of oxohistidine, hydroxytryptophan, and hydroxyaspartate in a plant protein has never been reported. Several examples of metal-catalyzed oxidation of His to oxohistidines were described for human proteins (Schöneich, 2000; Hovorka et al., 2002). Hydroxytryptophan is described as a precursor of the neurotransmitter serotonin and as an intermediary in tryptophan metabolism (Birdsall, 1998). Hydroxyaspartate residues were found in epidermal growth factor domains of several mammalian proteins. It was proposed that this modification regulates negatively the fucosylation of these regions (Castellino et al., 2008). In plants, oxidative modifications are known to occur in the presence of reactive oxygen species and metals, in particular in cell walls (Møller et al., 2007). In the case of At3g16850, suspected oxidized or hydroxylated amino acids do not belong to the active site according to bioinformatic prediction. We can note that in the closest plant homologs of At3g16850, the pentapeptide DPSHW is conserved, suggesting its importance for protein function. However, since no structural data are available and no catalytic residues have been identified for any plant polygalacturonase, it is still difficult to predict a role for these amino acids.
|
Identification and Characterization of Recalcitrant Structural CWPs
In all the plant cell wall proteomes described so far, structural cell wall glycoproteins such as extensins and H/PRPs represent only 1.7% of the presently identified CWPs and remain under-represented (Jamet et al., 2008). Many reasons may explain this situation: (1) the extraction of structural proteins is difficult, since they can be strongly bound to the cell wall; (2) the separation by electrophoresis in denaturing conditions can be disturbed by the presence of glycosylations; (3) digestion into peptides suitable for standard MS experiments can be limited because of the low frequency of protease cut sites into repetitive sequences, rendering protein identification by PMF difficult. Alternative strategies must be used to identify and characterize these proteins, such as peptide sequencing by tandem MS and MS in linear mode.
The At5g14920 H/PRP, also named AtGASA14 (Roxrud et al., 2007), was identified in etiolated hypocotyls and a smear was observed after 1D-E, with an apparent molecular mass of between 35 and 40 kDa (between K14 and K11 bands in Figure 6A) (Irshad et al., 2008). Aside from the Pro-rich region covering most of the protein, At5g14920 also contains a cleavable N-terminal signal peptide and a C-terminal sequence that shows homology with the GASA peptides previously described in plants (Figure 6B). Analysis of the primary sequence of the Pro-rich region indicates that it contains only two tryptic cut sites, since this protease cannot cleave Lys-Pro linkages. Altogether, in the entire sequence, only 12 Lys/Arg residues can be used for tryptic digestion (in bold in Figure 6B). In addition, some cleavage sites are very close together, generating peptides too small to be observed in the standard reflectron mode MALDI–TOF m/z window between 750 and 4000 Da. Instead, the two tryptic cut sites in the Pro-rich region release two large tryptic peptides, P1 and P2, which are also out of that m/z range. Despite these sequence constraints, three peptides from the C-terminal sequence, suitable for the standard reflectron mode MALDI–TOF MS, were found in the previously obtained spectra (Figure 6B). Their sequences were confirmed by MALDI–TOF–TOF MS/MS fragmentation (Supplemental Figure 6). In addition, N-terminal sequencing by Edman degradation (Irshad et al., 2008) enabled localizing the N-terminus of the protein downstream of the N-terminus predicted by bioinformatics, revealing a possible N-terminal processing or degradation. Thus, both tandem MS and Edman degradation results allowed correctly identifying At5g14920.
|
In order to find the P1 and P2 peptides in the Pro-rich region of At5g14920, we performed additional MALDI–TOF MS experiments, this time in linear mode instead of the reflectron mode used in PMF standard conditions. The presence of consecutive Pro in the Pro-rich region suggests that, similarly to other cell wall H/PRPs (Kieliszewski and Lamport, 1994), some of them may be hydroxylated. Up to 17 and 39 Hyp residues can be expected on P1 and P2, respectively, respectively increasing their mass up to 272 and 624 Da. This calculation excludes Hyp in the Leu–Pro and Lys–Pro contexts according to previous work (Kieliszewski and Lamport, 1994; Shimizu et al., 2005). The MS spectrum for K11 clearly shows two major peaks in the m/z range 5700–18 000 (Figure 6C). The peak centered on the mean m/z 15 400 possibly corresponds to P2. This m/z, which is higher than the peptide theoretical mass (between 11 859 and 12 483 Da, depending on the number of Hyp), suggests the presence of O-glycosylations in this region. In addition, the width of the observed peak (m/z 14 200–16 600) reflects mass heterogeneity, probably due to different states of O-glycosylation. The di-protonated form of P2 (m/z 7700) is also observed. However, the corresponding peak in the m/z range 7000–8800 may also contain the signal of the mono-protonated form of P1, which can overlap with that of the di-protonated P2. The spectrum obtained for K14 has a peak at m/z 13 070, namely within the m/z 12 500–14 000 range corresponding to P2. Interestingly, for this peptide, a shift of mass of about 2300 Da is observed between K11 and K14, indicating a lower level of O-glycosylation in K14. In K14, the di-protonated form of P2 is resolved from the mono-protonated form of P1, at m/z 6500. The peak centered on m/z 7200, namely within the m/z 7000–7500 range, may be attributed to the mono-protonated form of P1. This mass range coincides with the expected mass of P1 calculated in the N-terminus determined by Edman degradation (up to 6866 Da, Figure 6B).
Considering the primary sequence of At5g14920 and the known codes for hydroxylation of Pro and subsequent O-glycosylations, At5g14920 may be O-gycosylated in different ways. The presence of the Thr–Pro modules suggests the presence of arabinogalactans as in AGPs (Tan et al., 2003). However, the staining of the protein with the Yariv reagent was negative (Supplemental Figure 7). The presence of Lys–Pro–Pro–Thr motifs as in the maize THRGP (threonine-hydroxyproline-rich glycoprotein) suggests the presence of arabinosylated Hyp (Kieliszewski and Lamport, 1987; Kieliszewski et al., 1990). However, similar motifs are present in AtAGP31 and in NaPRP4, which were found to contain 80 and 83% of galactose, respectively (Sommer-Knudsen et al., 1996; Liu and Mehdy, 2007). Staining of a fraction enriched in At5g14920 with a lectin specific for galactose was positive, indicating the possible presence of galactose (Supplemental Figure 7). Finally, the presence of both arabinose and galactose cannot be excluded as for the Chlamydomonas reinhardtii GP1 HRGP (Ferris et al., 2001).
The MS results reported here suggest the presence of O-glycosylation on an H/PRP. As for NaPRP4 and AtAGP31 (Sommer-Knudsen et al., 1996; Liu and Mehdy, 2007), varying degrees of O-glycosylation were observed for At5g14920. The carbohydrate content of At5g14920 is estimated to be as much as 25% under that observed for NaPRP4 and AtAGP31 (75 and 80%, respectively), but much higher than that observed for soybean PRP2 (less than 1%) (Datta et al., 1989). A functional study of glycosylations should prove interesting, since the role of O-glycosylation of H/PRPs has so far never been investigated. In the case of extensins, O-glycosylation was shown to keep proteins in an extended conformation (Stafstrom and Staehelin, 1986). Further work is necessary to elucidate the structure of the Pro-rich region and of its glycosylations (Kieliszewski, 2001; Schultz et al., 2004; Tan et al., 2004; Shimizu et al., 2005). However, other biochemical approaches should be considered, since MS or MS/MS alone will not fully reveal the structural features of this recalcitrant protein, such as the location of Hyp residues.
Mass Spectrometry as a Diagnosis Tool for Protein Structure
The various examples presented in this article illustrate that MS is a powerful tool to gain new insight into the structure of CWPs in relation to their function. However, MS does not always allow the full coverage of a protein sequence, the complete description of PTMs, and the identification of all the proteins in a complex protein mixture. It appears that some proteins or peptides present in a sample fail to be analyzed. In certain cases, the MS results should be confirmed with additional experiments. N- or C-terminal sequencing can be used to determine the location of the ends of a protein, and consequently uphold biological processes such as maturation events. To go further into the three-dimensional structure of proteins, other methods can be used, such as nuclear magnetic resonance (NMR) or X-ray crystal diffraction (Liu and Hsu, 2005). However, the amount of purified proteins required for these analyses can be limiting as compared to MS. Chemical proteomics that consists of the analysis by MS of proteins specifically tagged at their site of modification is an emerging field that offers striking insights into the functional biology of PTMs (Tate, 2008). In conclusion, MS will increasingly be used as a diagnosis tool to address biological questions given its simple use, high throughput capacity, unequaled sensitivity, and the continuous technical developments in the field.
|
Methods |
|---|
|
TOPAbstractINTRODUCTIONRESULTS AND DISCUSSIONMethodsSUPPLEMENTARY DATAFUNDING |
|---|
Plant Material and Cell Wall Isolation
Arabidopsis thaliana seeds (ecotype Columbia 0) were grown in the dark as described (Feiz et al., 2006; Irshad et al., 2008), and etiolated hypocotyls were collected at 5 and 11 d. One to 1.8 cm hypocotyls were used to isolate cell walls as previously reported (Feiz et al., 2006). The obtained cell wall fraction was ground in liquid nitrogen in a mortar with a pestle prior to lyophilization.
Cell Wall Protein Extraction and Separation
Proteins were extracted from the cell wall preparation by successive steps of 0.2 M CaCl2 followed by 2 M LiCl. The proteins were separated by cation exchange chromatography (CEC). The column fractions were separated by 1D-electrophoresis (1D-E), stained and named as previously described (Irshad et al., 2008). The bands stained with Coomassie Brilliant blue were excised from the gels and digested with trypsin as previously described (Borderies et al., 2003; Boudart et al., 2005). Staining of proteins with the β-glycosyl Yariv reagent was performed as previously described (Willats and Knox, 1996). Staining of proteins with the galactose-specific peanut agglutinin (PNA) lectin was carried out using the DIG Glycan Differentiation Kit procedure (Roche, Mannheim, Germany).
Anhydrous Hydrogen Fluoride (HF) Deglycosylation
A sample of salt extractable CWP was HF-deglycosylated for 1 h at 4°C as described (Mort and Lamport, 1977; Shpak et al., 1999). The HF was blown off under nitrogen gas, and the deglycosylated proteins were then separated by 1D-E.
Mass Spectrometry (MS) Analyses
Sample preparation for all MALDI MS analyses was performed as previously described (Borderies et al., 2003). MALDI–TOF MS analyses were performed using a Voyager-DE STR mass spectrometer (Applied Biosystems/MDS, Sciex, USA). Spectra were acquired either in reflectron mode as previously reported (Borderies et al., 2003; Boudart et al., 2005) or in linear mode for a mass range of 2–20 kDa using the following parameters: accelerating voltage 25 kV, grid voltage 80% and acceleration delay time 800 ns. Every single acquisition run was composed of 2000 laser pulses at 20 Hz. Resolution of MALDI–TOF MS in linear mode was >300, corresponding to a full width at half maximum <50 for a peptide of m/z 15 000. The accuracy of the mass determination was ±0.4%, corresponding to m/z <80 in the mass region covered in linear mode.
MALDI–TOF–TOF analyses were performed using a MALDI–TOF–TOF Voyager 4700 (Applied Biosystems/MDS, Sciex, USA). MS and MS/MS data were recorded using the following parameters: accelerating voltage 8 and 15 kV (source 1 and source 2, respectively) and grid voltage 86%. The mass selection of the precursor ion was achieved using a mass window of ±5 Da and collision was performed in CID off mode. Data were acquired with 3750 shoots/spectrum.
Peptide mass fingerprints (PMFs) were compared to the non-redundant database of A. thaliana at NCBI (www.ncbi.nlm.nih.gov/) using ProteinProspector MS-FIT (http://prospector.ucsf.edu/prospector/cgi-bin/msform.cgi?form=msfitstandard). The criteria used for the database search were: a monoisotopic mass accuracy of 20 ppm; one missed cleavage; and hydroxylation of Pro residues. MS/MS results were analyzed using Protein Prospector MS-Product (http://prospector.ucsf.edu/prospector/cgi-bin/msform.cgi?form=msproduct).
Bioinformatics
Sub-cellular location, length of signal peptides, prediction of transmembrane domains, homologies to other proteins and protein functional domains were predicted as previously described (Minic et al., 2007) or using the ProtAnnDB database (www.polebio.scsv.ups-tlse.fr/ProtAnnDB/). ProtAnnDB collects predictions of sub-cellular location and of functional domains obtained with different software (San Clemente et al., 2009). N-glycosylation consensus sequences were determined using Prosite (www.expasy.org/prosite/). Functional domains were predicted by InterProScan (www.ebi.ac.uk/Tools/InterProScan/). All A. thaliana and O. sativa cell wall proteomic data are collected in the dedicated WallProtDB database (www.polebio.scsv.ups-tlse.fr/WallProtDB/).
The ProTerNyc software was developed in PHP5. It includes an N-glycan database associated with their molecular masses according to Faye et al. (2005). From a protein sequence and a mass list, ProTerNyc allows (1) prediction of a signal peptide using TargetP (www.cbs.dtu.dk/services/TargetP/), and (2) prediction of peptides possibly N-glycosylated using PROSITE (www.expasy.org/prosite/). In a first step, ProTerNyc predicts the length of the signal peptide. When the prediction is positive, a search for the presence of N-glycosylation motifs (PS00001) is carried out. The protein is cleaved following the prediction and submitted to MS-digest (http://prospector.ucsf.edu/prospector/cgi-bin/msform.cgi?form=msdigest). The list of masses obtained after theoretical digestion and eventually N-glycosylations on predicted consensus sites is compared to the list of experimental masses obtained after MALDI–TOF MS analysis with a mass tolerance of up to 60 ppm. The software accepts a tolerance for the length of the signal peptide around the position predicted by TargetP. The results are given in a file that can be exported in the Microsoft Office Excel format (www.microsoft.com/france/office/2007/programs/excel/overview.mspx).
|
SUPPLEMENTARY DATA |
|---|
|
TOPAbstractINTRODUCTIONRESULTS AND DISCUSSIONMethodsSUPPLEMENTARY DATAFUNDING |
|---|
Supplementary Data are available at Molecular Plant Online.
|
FUNDING |
|---|
|
TOPAbstractINTRODUCTIONRESULTS AND DISCUSSIONMethodsSUPPLEMENTARY DATAFUNDING |
|---|
This study was funded by the Université Paul Sabatier in Toulouse and the Centre National de la Recherche Scientifique (France).
|
Acknowledgements |
|---|
The authors are grateful to Prof. M. Kielieszewski (University of Ohio, USA) for welcoming GB in her laboratory for HF deglycosylation of protein samples, and to Drs J. Sampedro (University of Santiago de Compostela, Spain), Z. Minic (University of Saskatchewan, Canada), and J. Pelloux (University of Amiens, France) for stimulating discussions. Their gratitude also goes to G. Borderies and C. Pichereaux for technical assistance. All MS analyses were performed on the Plate-forme Protéomique Toulouse Midi-Pyrénées, France. The authors thank D. Mosovich for editing the English of the text. No conflict of interest declared.
-
Aebersold R, Mann M. Mass spectrometry-based proteomics. Nature (2003) 422:198–207.[CrossRef][Medline]
Alvarez S, Goodger JQ, Marsh EL, Chen S, Asirvatham VS, Schachtman DP. Characterization of the maize xylem sap proteome. J. Proteome Res. (2006) 5:963–972.[CrossRef][Web of Science][Medline]
Baumberger N, et al. Whole-genome comparison of leucine-rich repeat extensins in Arabidopsis and rice: a conserved family of cell wall proteins form a vegetative and a reproductive clade. Plant Physiol (2003) 131:1313–1326.
Bayer EM, Bottrill AR, Walshaw J, Vigouroux M, Naldrett MJ, Thomas CL, Maule AJ. Arabidopsis cell wall proteome defined using multidimensional protein identification technology. Proteomics (2006) 6:301–311.[CrossRef][Web of Science][Medline]
Berger D, Altmann T. A subtilisin-like serine protease involved in the regulation of stomatal density and distribution in Arabidopsis thaliana. Genes Dev. (2000) 14:1119–1131.
Bhushan D, Pandey A, Chattopadhyay A, Choudhary MK, Chakraborty S, Datta A, Chakraborty N. Extracellular matrix proteome of chickpea (Cicer arietinum L.) illustrates pathway abundance, novel protein functions and evolutionary perspect. J. Proteome Res. (2006) 5:1711–1720.[CrossRef][Web of Science][Medline]
Birdsall T. 5-Hydroxytryptophane: a clinically-effective serotonin precursor. Altern. Med. Rev. (1998) 3:271–280.
Borderies G, Jamet E, Lafitte C, Rossignol M, Jauneau A, Boudart G, Monsarrat B, Esquerré-Tugayé MT, Boudet A, Pont-Lezica R. Proteomics of loosely bound cell wall proteins of Arabidopsis thaliana cell suspension cultures: a critical analysis. Electrophoresis (2003) 24:3421–3432.[CrossRef][Web of Science][Medline]
Bosch M, Cheung A, Hepler P. Pectin methylesterase, a regulator of pollen tube growth. Plant Physiol (2005) 138:1334–1346.
Boudart G, Jamet E, Rossignol M, Lafitte C, Borderies G, Jauneau A, Esquerré-Tugayé M-T, Pont-Lezica R. Cell wall proteins in apoplastic fluids of Arabidopsis thaliana rosettes: identification by mass spectrometry and bioinformatics. Proteomics (2005) 5:212–221.[CrossRef][Web of Science][Medline]
Carpita NC, Gibeaut DM. Structural models of primary cell walls in flowering plants: consistency of molecular structure with the physical properties of the walls during growth. Plant J (1993) 3:1–30.[CrossRef][Web of Science][Medline]
Caspers M, Lok F, Sinjorgo K, van Zeijl M, Nielsen K, Cameron-Mills V. Synthesis, processing and export of cytoplasmic endo-β-1,4-xylanase from barley aleurone during germination. Plant J (2001) 26:191–204.[CrossRef][Web of Science][Medline]
Castellino F, Ploplis V, Zhang L. 
-glutamate and β-hydroxyaspartate in proteins. Methods Mol. Biol. (2008) 446:85–94.[CrossRef][Medline]
Charmont S, Jamet E, Pont-Lezica R, Canut H. Proteomic analysis of secreted proteins from Arabidopsis thaliana seedlings: improved recovery following removal of phenolic compounds. Phytochemistry (2005) 66:453–461.[CrossRef][Web of Science][Medline]
Chen X, Kim S, Cho W, Rim Y, Kim S, Kim S, Kang K, Park Z, Kim J. Proteomics of weakly bound cell wall proteins in rice calli. J. Plant Physiol (2008) 166:665–685.
Chevalier F, Rofidal V, Vanova P, Bergoin A, Rossignol M. Proteomic capacity of recent fluorescent dyes for protein staining. Phytochemistry (2004) 65:1499–1506.[CrossRef][Web of Science][Medline]
Cho W, Chen X, Chu H, Rim Y, Kim S, Kim S, Kim S-W, Park Z-Y, Kim J-Y. The proteomic analysis of the secretome of rice calli. Physiol. Plant (2009) 135:331–341.
Datta K, Schmidt A, Marcus A. Characterization of two soybean repetitive proline-rich proteins and a cognate cDNA from germinated axes. Plant Cell. (1989) 1:945–952.
Ellis C, Karafyllidis I, Wasternack C, Turner J. The Arabidopsis mutant cev1 links cell wall signaling to jasmonate and ethylene responses. Plant Cell. (2002) 14:1557–1566.
Emanuelsson O, Brunak S, Von Heijne G, Nielsen H. Locating proteins in the cell using TargetP, SignalP and related tools. Nat. Protoc (2007) 2:953–971.[CrossRef]
Faye L, Boulaflous A, Benchabane M, Gomord V, Michaud D. Protein modifications in the plant secretory pathway: current status and practical implications in molecular pharming. Vaccine (2005) 23:1770–1778.[CrossRef][Web of Science][Medline]
Feiz L, Irshad M, Pont-Lezica RF, Canut H, Jamet E. Evaluation of cell wall preparations for proteomics: a new procedure for purifying cell walls from Arabidopsis hypocotyls. Plant Methods (2006) 2:10.[CrossRef][Medline]
Ferris P, Woessner J, Waffenschmidt S, Kilz S, Drees J, Goodenough U. Glycosylated polyproline II rods with kinks as a structural motif in plant hydroxyproline-rich glycoproteins. Biochemistry (2001) 40:2978–2987.[CrossRef][Web of Science][Medline]
Fry SC. Primary cell wall metabolism: tracking the careers of wall polymers in living plant cells. New Phytol (2004) 161:641–675.[CrossRef][Web of Science]
Gray J. Heterogeneity of glycans at each N-glycosylation site of horseradish peroxidase. Carbohydr. Res. (1998) 311:61–69.[CrossRef][Web of Science][Medline]
Gray J, Montgomery R. Asymmetric glycosylation of soybean seed coat peroxidase. Carbohydr. Res. (2006) 341:198–209.[CrossRef][Web of Science][Medline]
Han X, Aslanian A, Yates J III. Mass spectrometry for proteomics. Curr. Opin. Chem. Biol. (2008) 12:483–490.[CrossRef]
Hovorka S, Biesiada H, Williams T, Hühmer A, Schöneich C. High sensitivity of Zn2+ insulin to metal-catalyzed oxidation: detection of 2-oxo-histidine by tandem mass spectrometry. Pharm. Res. (2002) 19:530–537.[CrossRef][Web of Science][Medline]
Irshad M, Canut H, Borderies G, Pont-Lezica R, Jamet E. A new picture of cell wall protein dynamics in elongating cells of Arabidopsis: confirmed actors and newcomers. BMC Plant Biol. (2008) 8:94.[CrossRef][Medline]
Ito Y, Nakanomyo I, Motose H, Iwamoto K, Sawa S, Dohmae N, Fukuda H. Dodeca-CLE peptides as suppressors of plant stem differentiation. Science (2006) 313:8842–8845.
Jamet E, Albenne C, Boudart G, Irshad M, Canut H, Pont-Lezica R. Recent advances in plant cell wall proteomics. Proteomics (2008) 8:893–908.[CrossRef][Web of Science][Medline]
Jamet E, Canut H, Boudart G, Pont-Lezica R. Cell wall proteins: a new insight through proteomics. Trends Plant Sci. (2006) 11:33–39.[CrossRef][Web of Science][Medline]
Jung Y-H, Jeong S-H, Kim S, Singh R, Lee J-E, Cho Y-S, Agrawal G, Rakwal R, Jwa N-S. Systematic secretome analyses of rice leaf and seed callus suspension-cultured cells: workflow development and establishment of high-density two-dimensional gel reference maps. J. Proteome Res. (2008) 7:5187–5210.[CrossRef][Web of Science][Medline]
Kieliszewski M. The latest hype on Hyp-O-glycosylation codes. Phytochemistry (2001) 57:319–323.[CrossRef][Web of Science][Medline]
Kieliszewski M, Lamport D. Purification and partial characterization of a hydroxyproline-rich glycoprotein in a graminaceous monocot, Zea mays. Plant Physiol (1987) 85:823–827.
Kieliszewski M, Lamport D. Extensin: repetitive motifs, functional sites, post-translational codes, and phylogeny. Plant J (1994) 5:157–172.[CrossRef][Web of Science][Medline]
Kieliszewski M, Leykam J, Lamport D. Structure of the threonine-rich extensin from Zea mays. Plant Physiol (1990) 92:316–326.
Kimura Y, Hess D, Sturm A. The N-glycans of jack bean -mannosidase. Eur. J. Biochem (1999) 264:168–175.[Web of Science][Medline]
Kondo T, Sawa S, Kinoshita A, Mizuno S, Kakimoto T, Fukuda H, Sakagami Y. A plant peptide encoded by CLV3 identified by in situ MALDI–TOF MS analysis. Science (2006) 313:845–848.
Kurdyukov S, et al. The epidermis-specific extracellular BODYGUARD controls cuticle development and morphogenesis in Arabidopsis. Plant Cell. (2006) 18:321–339.
Lamport D. The protein component of primary cell walls. Adv. Bot. Res. (1965) 2:151–218.
Lee R, Hrmova M, Burton R, Lahnstein J, Fincher G. Bifunctional family 3 glycoside hydrolases from barley with -L-arabinofuranosidase and β-D-xylosidase activity: characterization, primary structures, and COOH-terminal processing. J. Biol. Chem. (2003) 278:5377–5387.
Lige B, Shengwu M, van Huystee R. The effects of the site-directed removal of N-glycosylation from cationic peanut peroxidase on its function. Arch. Biochem. Biophys (2001) 386:17–24.[CrossRef][Web of Science][Medline]
Lim J-M, Aoki K, Angel P, Garrison D, King D, Tiemeyer M, Bergmann C, Wells L. Mapping glycans onto specific N-linked glycosylation sites of Pyrus communis PGIP redefines the interface for EPG–PGIP interactions. J. Proteome Res. (2009) 8:673–680.[CrossRef][Web of Science][Medline]
Liu C, Mehdy M. A nonclassical arabinogalactan protein gene highly expressed in vascular tissues, AGP31, is transcriptionally repressed by methyl jasmonic acid in Arabidopsis. Plant Physiol (2007) 145:863–874.
Liu H-L, Hsu J-P. Recent development in structural proteomics for protein structure determination. Proteomics (2005) 5:2056–2068.[CrossRef][Web of Science][Medline]
Micheli F. Pectin methylesterases: cell wall enzymes with important roles in plant physiology. Trends Plant Sci. (2001) 6:414–419.[CrossRef][Web of Science][Medline]
Minic Z, Jamet E, Negroni L, der Garabedian PA, Zivy M, Jouanin L. A sub-proteome of Arabidopsis thaliana trapped on Concanavalin A is enriched in cell wall glycoside hydrolases. J. Exp. Bot (2007) 58:2503–2512.
Minic Z, Rihouey C, Do C, Lerouge P, Jouanin L. Purification and characterization of enzymes exhibiting β-D-xylosidase activities in stem tissues of Arabidopsis. Plant Physiol (2004) 135:867–878.
Møller I, Jensen P, Hansson A. Oxidative modifications to cellular components in plants. Ann. Rev. Plant Biol. (2007) 58:459–481.[CrossRef]
Mort AJ, Lamport DT. Anhydrous hydrogen fluoride deglycosylates glycoproteins. Anal. Biochem (1977) 82:289–309.
Oliva M, Theiler G, Zamocky M, Koua D, Margis-Pinheiro M, Passardi F, Dunand C. PeroxiBase: a powerful tool to collect and analyse peroxidase sequences from Viridiplantae. J. Exp. Bot (2009) 60:453–459.
Pattison R, Amtmann A. N-glycan production in the endoplasmic reticulum of plants. Trends Plant Sci. (2009) 14:92–99.[CrossRef][Web of Science][Medline]
Roberts K. Structures at the plant cell surface. Curr. Opin. Cell Biol. (1990) 2:920–928.
Roberts K. How the cell wall acquired a cellular context. Plant Physiol (2001) 125:127–130.
Rose J, Braam J, Fry S, Nishitani K. The XTH family of enzymes involved in xyloglucan endotransglucosylation and endohydrolysis: current perspectives and a new unifying nomenclature. Plant Cell Physiol (2002) 43:1421–1435.
Roudier F, et al. COBRA, an Arabidopsis extracellular glycosyl-phosphatidyl inositol-anchored protein, specifically controls highly anisotropic expansion through its involvement in cellulose microfibril orientation. Plant Cell. (2005) 17:1749–1763.
Roxrud I, Lid S, Fletcher J, Schmidt E, Opsahl-Sorteberg H. GASA4, one of the 14-member Arabidopsis GASA family of small polypeptides, regulates flowering and seed development. Plant Cell Physiol (2007) 48:471–483.
Sampedro J, Sieiro C, Revilla G, González-Villa T, Zarra I. Cloning and expression pattern of a gene encoding an -xylosidase active against xyloglucan oligosaccharides from Arabidopsis. Plant Physiol (2001) 126:910–920.
San Clemente H, Pont-Lezica R, Jamet E. Bioinformatics as a tool for assessing the quality of sub-cellular proteomic strategies and inferring functions of proteins: plant cell wall proteomics as a test case. Bioinform. Biol. Insights (2009) 3:15–28.
Schöneich C. Mechanisms of metal-catalyzed oxidation of histidine to 2-oxo-histidine in peptides and proteins. J. Pharm. Biomed. Anal (2000) 21:1093–1097.[CrossRef][Web of Science][Medline]
Schultz CJ, Ferguson KL, Lahnstein J, Bacic A. Post-translational modifications of arabinogalactan-peptides of Arabidopsis thaliana. J. Biol. Chem. (2004) 279:455103–445511.
Shimizu M, Igasaki T, Yamada M, Yuasa K, Hasegawa J, Kato T, Tsukagoshi H, Nakamura K, Fukuda H, Matsuoka K. Experimental determination of proline hydroxylation and hydroxyproline arabinogalactosylation motifs in secretory proteins. Plant J (2005) 42:877–889.[CrossRef][Web of Science][Medline]
Showalter A. Structure and function of plant cell wall proteins. Plant Cell. (1993) 5:9–23.
Shpak E, Leykam J, Kieliszewski M. Synthetic genes for glycoprotein design and the elucidation of hydroxyproline-O-glycosylation codes. Proc. Natl Acad. Sci. U S A (1999) 21:14736–14741.
Somerville C, et al. Toward a systems approach to understanding plant cell walls. Science (2004) 306:2206–2211.
Sommer-Knudsen J, Clarke A, Bacic A. A galactose-rich, cell-wall glycoprotein from styles of Nicotiana alata. Plant J (1996) 9:71–83.[CrossRef][Web of Science][Medline]
Stafstrom JP, Staehelin LA. The role of carbohydrate in maintaining extensin in an extended conformation. Plant Physiol (1986) 81:242–246.
Tan L, Leykam JF, Kieliszewski MJ. Glycosylation motifs that direct arabinogalactan addition to arabinogalactan-proteins. Plant Physiol (2003) 132:1362–1369.
Tan L, Qiu F, Lamport D, Kieliszewski M. Structure of a hydroxyproline (Hyp)-arabinogalactan polysaccharide from repetitive Ala-Hyp expressed in transgenic Nicotiana tabacum. J. Biol. Chem. (2004) 279:13156–13165.
Tate E. Recent advances in chemical proteomics: exploring the post-translational proteome. J. Chem. Biol. (2008) 1:17–26.[CrossRef]
Vogel J, Raab T, Somerville C, Somerville S. Mutations in PMR5 result in powdery mildew resistance and altered cell wall composition. Plant J (2004) 40:968–978.[CrossRef][Web of Science][Medline]
Watson BS, Lei Z, Dixon RA, Sumner LW. Proteomics of Medicago sativa cell walls. Phytochemistry (2004) 65:1709–1720.[CrossRef][Web of Science][Medline]
Willats WG, Knox JP. A role for arabinogalactan-proteins in plant cell expansion: evidence from studies on the interaction of β-glucosyl Yariv reagent with seedlings of Arabidopsis thaliana. Plant J (1996) 9:919–925.[CrossRef][Web of Science][Medline]
Willats WG, McCartney L, Mackie W, Knox JP. Pectin: cell biology and prospects for functional analysis. Plant Mol. Biol. (2001) 47:9–27.[CrossRef][Web of Science][Medline]
Wolf S, Rausch T, Greiner S. The N-terminal pro region mediates retention of unprocessed type-I PME in the Golgi apparatus. Plant J (2009) 58:361–375.[CrossRef][Web of Science][Medline]
Zhu J, Chen S, Alvarez S, Asirvatham VS, Schachtman DP, Wu Y, Sharp RE. Cell wall proteome in the maize primary root elongation zone. I. Extraction and identification of water-soluble and lightly ionically bound proteins. Plant Physiol (2006) 140:311–325.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||