Journal of Experimental Botany, Vol. 54, No. 384, pp. 971-983,
March 1, 2003
© 2003 Oxford University Press
A Tudor protein with multiple SNc domains from pea seedlings: cellular localization, partial characterization, sequence analysis, and phylogenetic relationships
Received 7 July 2002; Accepted 25 November 2002
1 Laboratory of Molecular Cell Biology, Department of Biological Resources, Faculty of Agriculture, Ehime University, Matsuyama, 790-8566, Japan
2 Botany Department, Box 7612, North Carolina State University, Raleigh NC 27695, USA
3 Present address: Laboratory of Cell Genetics, Research Institute for Bioresources, Okayama University, Kurashiki, Okayama 710-0046, Japan.
4 To whom correspondence should be addressed: Fax: +81 89 946–9853. e-mail: abe{at}mcb.agr.ehime-u.ac.jp
Abbreviations: CDB, cytoskeleton depolymerizing buffer; CSB, cytoskeleton stabilizing buffer; DTT, dithiothreitol; EGTA, ethylenebis(oxyethylenenitrilo)tetra-acetic acid; HEPES, N-2-hydroxyethylpiperazine-N'-2-ethanesulphonic acid; BPB, bromophenol blue; IEF, isoelectric focusing; PMSF, phenylmethylsulphonyl fluoride; PTE, polyoxyethylene-10-tridecyl ether; SDS, sodium dodecyl sulphate; SDS-PAGE, SDS-polyacrylamide gel electrophoresis; Triton X-100, polyethylene glycol mono-p-iso-octylphenyl ether; TRIS, tris-(hydroxymethyl) aminomethane.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionReferences |
|---|
A major high molecular weight protein (HMP) in the cytoskeletal fraction from pea has been purified. A combination of chromatographic techniques and protease fragment analysis also facilitated the isolation of the encoding cDNA, disclosing the sequence of the complete open reading frame. The protein possesses four complete N-terminal Staphylococcal nuclease (SNc) domains, a central Tudor domain and a partial SNc domain at the C-terminus, which may act as a coiled-coil cytoskeleton interaction motif. Cell fractionation studies showed that the protein was abundant in the cytoskeleton fraction in dark-grown pea seedlings, but essentially was absent from the nucleus. Gel filtration column chromatography indicated that the native protein exists as a dimer, while isoelectric focusing suggested that there were at least four HMP isotypes. The protein co-eluted with ribosomes from a heparin affinity column in vitro, consistent with ribosome/polysome interactions in vivo. Significantly, sequence analysis of the C-terminal SNc motif may accurately predict nuclear versus cytoplasmic localization resulting in potentially very different functional roles for this protein family in different organisms. An antibody to HMP from peas was also raised and an HMP with a similar molecular mass was detected in the cytoskeleton fractions and to a lesser extent in the nuclear fraction (250 g pellet) from rice and wheat seedlings.
Key words: Cytoskeleton, heparin, IEF, localization, p100, PAGE, phylogeny, SNc, Tudor, V8 protease.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
The cytoskeleton exerts multiple functions in plant cells including the targeting, tethering, transport, and translation of mRNA (Davies et al., 1996, 1998, 2001). In previous work, various proteins present in the cytoskeleton fraction were isolated and identified and attention was focused on those likely to be involved in ribosome and mRNA localization. Techniques were developed for isolating the cytoskeleton–membrane–polysome complex in reasonably pure form, then removing the membranes with non-ionic detergents, depolymerizing the cytoskeleton, removing the polysomes by ultra-centrifugation, and analysing the released (putative cytoskeleton) proteins (Abe et al., 1992, 1994; Ito et al., 1994; Abe and Davies, 1995). More than 15 distinct proteins in the cytoskeleton fraction have been described (Davies et al., 2001), one of which migrated as a 110 kDa protein in SDS-PAGE (Abe and Davies, 1991; Abe et al., 1992; Ito et al., 1994). Solubilized cytoskeleton proteins re-aggregate when applied to columns, but this is prevented by heparin (Shibata et al., 1999), and so heparin affinity column chromatography was developed for their initial separation. Using this technique, it was shown that this 110 kDa protein (HMP) eluted from the column at about 390 mM KOAc and was one of the major proteins found in the cytoskeleton fraction in dark-grown pea seedlings (Shibata et al., 1999). The HMP was then identified as a cytoskeleton-bound Tudor protein having multiple SNc domains (DDBJ/EMBL/GenBank Accession No. AB055904). A similar, 120 kDa Tudor protein with four SNc domains has been identified in developing rice endosperm cells (Sami-Subbu et al., 2001) and was thought to be homologous to the 100 kDa human Tudor protein. The 120 kDa Tudor protein from rice has high affinity to RNA and possesses a coiled-coil motif, which may bind to the cytoskeleton (Sami-Subbu et al., 2001). Information also exists in databases on genomic fragments encoding similar proteins in Arabidopsis thaliana (CAB87924, BAB10078), but the function and subcellular localization have not yet been identified experimentally.
The Staphylococcal nuclease (SNc) domain is present in many eukaryotic proteins and has been implicated in binding to RNA or single-stranded DNA (Tong et al., 1995), but these SNc domains in p100 lack the nucleolytic active site residues of Staphylococcal nuclease (Keefe et al., 1993, 1994). In Drosophila and other species, there are several SNc domain proteins, which also possess a distinct region called a Tudor domain, whose exact function is unknown, but which can bind RNA and, in some cases, DNA (Ponting, 1997). A unique 100 kDa Tudor domain protein containing four SNc domains has been found to be associated with infection by the Epstein Barr virus in humans (Tong et al., 1995) and a pathogenic fungus in mouse (Porta et al., 1999), and was originally thought to function as a transcription co-factor associated with pathogenic invasion. Recently, however, expression of the human p100 has been reported in mammary gland cells (Broadhurst and Wheeler, 2001), and thus the protein seems to have a function in normal cells. This group of apparently distinct 100–120 kDa Tudor proteins with four SNc domains is of special interest, in part, because their high degree of conservation implies an essential, indispensable role in the function of normal cells.
In the present study, a 110 kDa protein (HMP) from the cytoskeleton of pea stem cells was isolated and purified using a heparin affinity column followed by gel filtration, and its partial amino acid sequence analysed. The cDNA encoding HMP was cloned to determine its complete nucleotide and amino acid sequence and protein structure. It is shown that HMP is a Tudor domain protein with five SNc domains. In plants (pea, rice, wheat), the majority of this Tudor protein is present in the cytoskeleton fraction and very little is in the nucleus, which strongly implies that it is not functioning solely as a transcriptional cofactor in plants, as it appears to be in animals. Finally, the phylogenetic relationships between homologues from a broad range of eukaryotes have been analysed that show distinct plant, mammalian, and fungal counterparts.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Plant material and fractionation
Pea (Pisum sativum L. var. Alaska) plants were grown in the dark for 5 d at 20–24 °C and the apical 10 mm of the first internode harvested. Tissue was ground with a mortar and pestle in 7 vols of a specially designed cytoskeleton-stabilizing buffer (CSB) consisting of 5 mM HEPES-KOH (pH 7.5), 10 mM Mg(OAc)2, 2 mM EGTA, and 1 mM PMSF with the addition of 0.5% polyoxyethylene-10-tridecyl ether (PTE). The homogenate was filtered through two layers of Miracloth (Calbiochem), centrifuged at 250 g for 5 min to remove nuclei and wall debris, the pellet generally discarded, and the supernatant centrifuged at 27 000 g to furnish the cytoskeleton pellet [1, 4]. The cytoskeleton pellet was resuspended in cytoskeleton-depolymerizing buffer (CDB) consisting of 200 mM TRIS-HCl (pH 8.5), 450 mM KOAc, 25 mM Mg(OAc)2, 2% PTE, and 0.5 mM ATP using a teflon homogenizer, and centrifuged at 27 000 g for 15 min to remove undissolved materials (Abe and Davies, 1995). The supernatant was centrifuged at 300 000 g for 1 h (L8-60M, Beckman; rotor 50.2 Ti) to remove ribosomes leaving behind the solubilized cytoskeleton-associated proteins.
Purification of HMP using column chromatography
Purification of HMP by heparin affinity column chromatography was performed as described previously (Shibata et al., 1999). Extracted proteins from various fractions were filtered through a 0.45 µm membrane (DISMIC-25cs, Toyo Roshi), diluted with 3 vols of dH2O, and applied immediately to a heparin affinity column (HiTrap Heparin, 1 ml bed volume, Amersham Pharmacia Biotech). The column was washed with 10 vols of CSB containing 0.15 M KOAc, eluted with a gradient of 0.15–1.0 M KOAc in CSB with the SMART System (Amersham Pharmacia Biotech), and washed with 10 vols 3 mg ml–1 heparin sodium salt in 5 mM HEPES-KOH (pH 7.5). Since the flow cell in the SMART SYSTEM tended to clog, the absorbance of an aliquot of each 300 µl gradient fraction was measured at 260, 280, and 320 nm using a spectrophotometer (DU-65, Beckmann). Fractions containing HMP were pooled, frozen in liquid N2 and stored at –85 °C. A gel filtration column (Superose, Amersham Pharmacia Biotech) was equilibrated with a buffer consisting of 400 mM KOAc, 5 mM HEPES-KOH (pH 7.5), 10 mM Mg(OAc)2, and 2 mM EGTA using the SMART System (Amersham Pharmacia Biotech). Then an 80 µl aliquot of stored HMP was injected, the column was eluted at a flow rate of 100 µl min–1 with the same buffer used in the equilibration, and the absorbance was continuously monitored at 260, 280, and 320 nm using a flow cell with 5 mm light path in the SMART SYSTEM, and 80 µl fractions were collected.
SDS-PAGE and 2D-PAGE
For SDS-PAGE, proteins were precipitated overnight with 4 vols of acetone and dissolved in a sample buffer containing 2% LDS and 0.01 M TRIS-HCl (pH 6.8), 20% glycerol, 1% ß-mercaptoethanol, 0.01% BPB, heated at 95 °C for 5 min, and then separated by electrophoresis. For 2D-PAGE, the acetone precipitate (above) was washed successively with 80% and 100% acetone, dissolved in 8 M urea, 0.5% NP-40, 2% ß-mercaptoethanol, 0.8% Pharmalyte® 4–7 for IEF (Amersham Pharmacia Biotech), and 0.01% BPB, and applied onto a drystrip gel (pI 4–7, Amersham Pharmacia Biotech) equilibrated with 8 M urea, 0.5% Triton-X100, 10 mM DTT, 2 mM acetic acid, and 0.01 mg ml–1 orange G and subject to IEF at 32 500 V h–1 using a Multiphor II 2D-Electrophoresis System (Amersham Pharmacia Biotech). The strip was immersed in 6 M urea, 0.05 M TRIS-HCl (pH 6.8), 30% glycerol, 1% SDS, and 16 mM DTT for 10 min, followed by 6 M urea, 0.05 M TRIS-HCl (pH 6.8), 30% glycerol, 1% SDS, and 0.01% BPB for 10 min, both with gentle shaking. The strip was then placed onto a separating slab gel for SDS-PAGE, electrophoresed in the second dimension, and proteins detected by staining with Brilliant Blue R250.
Determination of partial amino acid sequences of HMP digested by V8 protease
The heparin affinity column fractions (Nos 20–30, Fig. 1B) or gel filtration column fractions (Nos 9–10, Fig. 2B) containing HMP were pooled and separated by SDS–PAGE. Bands of HMP stained with Brilliant Blue R-250 were cut out, loaded on a stacking gel, and the protein was digested with Staphylococcus aureus V8 protease (EC 3.4.21.19) as described by Cleveland et al. (1977). The digested polypeptides were separated on 15% acrylamide gels using SDS–PAGE and electroblotted onto a PVDF membrane (Immobilon-P, Millipore). The transferred peptides were detected by staining with Brilliant Blue R250, the peptide bands were then excised and sequenced using a protein sequencer (Model 476A, Perkin Elmer Applied Biosystems Division).
|
|
Isolation of intact HMP cDNA
Total RNA was isolated from similar tisssue to that described in Plant materials and fractionation, according to a method described earlier (Shibata et al., 2001), and mRNA was obtained by using an mRNA purification kit (Amersham Pharmacia biotech) from the total RNA was used for synthesizing cDNA. To generate a fragment of cDNA encoding HMP using PCR, the forward primer, 5'-AGAGATGTTCGCATTGTTCTTGA-3' encoding fragment V8-5 and the reverse primer, 3'-TACCTAAAACCGAGTCACAA-5' encoding fragment CNBr2 were employed. The fragment obtained was subcloned in pMOSBlue vecter using pMOSBlue vector blunt-ended cloning kit (Amersham Pharmacia Biotech) and sequenced. cDNA for 3' RACE was synthesized using a tagged oligo-dT primer 5'-aagaattctcgagctccagaa-t25-3'. 3' RACE was performed with a forward primer 5'-gtccagttgatgcagctgga-3' made on the sequence obtained and the reverse primer made on the tagged oligo-dT primer 5'-aagaattctcgagctccag-3' for amplification of the 3' region, a fragment of 1825 bp, subcloned, sequenced, and shown to be 2343 bp (No. AB055904). cDNA was synthesized by MMLV reverse transcriptase (ReverTra Ace, Toyobo, Japan) with a reverse primer, 3'-tcgacgaggattg cccattt-5', the blunt-ended second strand was synthesized using simultaneous inclusion of DNA polymerase I (9 U µl–1), E. coli DNA ligase (6 U µl–1, Takara, Japan), and T4 DNA polymerase (9 U µl–1) in the presence of RNase H (1 U µl–1) in a reaction buffer composed of 33.4 mM TRIS-HCl (pH 8.5), 4.6 mM MgCl2, 10 mM (NH4)2SO4, 6.6 mM ß-mercaptoethanol, 0.06 mM EDTA, 0.005% BSA, 100 mM KCl, 0.15 mM ß-NAD, and ligated to an adaptor consisting of complementary oligo DNA, 5'-gtaatacgactcactataggg cacgcgtggtcgacggcccgggctggt-3' and 3'-gggccc gacca-5', using 4 U µl–1 T4 DNA ligase. T4 DNA ligase, DNA polymerase I, and RNase H were obtained from Toyobo, Japan. Using this ligated product, 5' RACE was performed with a primer which annealed to the adaptor, 5'-gtaatacgactcactatag ggc-3', and a gene specific reverse primer made on the sequence in AB055904, 3'-CCTCAACCAACTTGCCAGTT-5'.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Isolation and purification of HMP from the cytoskeleton fraction
Pea epicotyl tissue was homogenized in cytoskeleton-stabilizing buffer (CSB), filtered through Miracloth, and centrifuged to furnish the crude nuclear (250 g) pellet and the crude cytoskeleton (27 000 g) pellet. The cytoskeleton was resuspended in cytoskeleton-depolymerizing buffer (CDB), and centrifuged at 27 000 g to furnish a supernatant containing polysomes and solubilized cytoskeleton proteins, and the supernatant was further centrifuged at 300 000 g to furnish the polysome pellet and the released (cytoskeleton) proteins. Proteins in these subcellular fractions were separated by SDS-PAGE, stained with Brilliant Blue R250, and shown in Fig. 1A.
In the 27 000 g pellet (cytoskeleton, Fig. 1A, lane e), there was a significant amount of a 110 kDa protein (HMP, indicated by the arrow) compared with the total homogenate (lane a) and the 250 g supernatant (lane c) or the 27 000 g supernatant (lane d). Virtually no Brilliant blue-stained band that corresponded to HMP was present in the nuclear pellet (lane b), while nuclear histones were abundant (asterisks in lane b), indicating this fraction was highly enriched for nuclei. After removing the ribosomes from the solubilized cytoskeleton pellet by centrifuging at 300 000 g for 1 h, HMP remained in the supernatant (lane f).
The 300 000 g supernatant of the cytoskeleton fraction depolymerized in CDB (Fig. 1A, lane f) was diluted with water and applied to a heparin affinity column equilibrated with 150 mM KOAc in CSB. After washing the column with 10 vols of 150 mM KOAc in CSB, proteins were eluted by a linear gradient of 0.15 M to 1 M KOAc in CSB, monitored at 260 nm and 280 nm, fractions were collected (Fig. 1B), and each of these 30 fractions was analysed by SDS-PAGE to determine their protein pattern (Fig. 1C). HMP was eluted at 380–420 mM KOAc (Fig. 1B, C, lanes 9–13), while a 49 kDa protein (apyrase) was eluted at about 640 mM KOAc (Fig. 1C, lanes 18–22), as previously reported (Shibata et al., 1999). Another RNA-interacting protein, RSP (ribosome-sedimenting protein) was in lanes 20–23 (Abe et al., 1995) and another protein, LMP (low molecular weight protein), was in lanes 19–24 (Fig. 1C).
Since the HMP isolated here contained small amounts of other proteins (Fig. 1C, lanes 9–13) and large amounts of 260 nm absorbing materials (Fig. 1B, fractions 9–13), it was purified further by gel filtration using an S12 superose column (Amersham Pharmacia Biotech), and its elution position compared with those of standard proteins (Fig. 2A). The 280 nm absorbance profile for the 300 000 g supernatant shows a major peak at fraction 9–10, corresponding to a size of 220 kDa, and a minor peak in the void volume at 2000 kDa and a very small hump at about 100 kDa (Fig. 2B). The profile for the total 27 000 g (cytoskeleton) pellet (i.e. the solubilized cytoskeleton proteins along with the retained ribosomes) shows a much bigger peak in the void volume and a smaller one at fraction 9–10 (Fig. 2C), while the corresponding A260 profile (indicating the presence of nucleic acids) shows even greater absorbance in the void volume peak and even less in the fraction 9–10 peak. This void volume fraction contained ribosomal proteins and ribosomal RNA (data not shown).
The peak fractions (Fig. 2B, fractions 9–10) containing 5 µg protein (overloaded to detect impurities) were subjected to SDS-PAGE, stained with Brilliant Blue R250, and the results shown in Fig. 3. A major band of 110 kDa HMP along with a few minor bands were observed in the gel filtrate (Fig. 3, lane ‘GF’). The protein pattern of the void volume from the cytoskeleton-ribosome fraction (Fig. 2C) was also analysed (Fig. 3B, lane ‘Vd’) and compared with purified ribosomes (Fig. 3B, lane ‘Rb’). Since these patterns are almost identical, it would appear that HMP associates with ribosomes or ribosomal proteins. When a sample identical to that analysed by 1D-PAGE (Fig. 3A, lane ‘GF’) was analysed by 2D-PAGE, at least four isotypes with pIs of 6.04, 6.13, 6.24, and 6.34 were seen (Fig. 3C).
|
Analysis of partial amino acid sequences of proteolytic fragments of HMP
When purified HMP (Fig. 3A, C) was desalted for further analysis, it became highly aggregated in dilute buffers. Accordingly, the HMP purified by gel filtration was separated by SDS–PAGE and the band corresponding to HMP was excised and subjected to digestion by V8 protease for varying times, to get variously sized fragments (Cleveland et al., 1977) and by CNBr using 10 mg ml–1 cyanogen bromide in 70% formic acid (Jahnen-Dechent and Simpson, 1990). After digestion, SDS–PAGE was performed to separate the peptide fragments, which were then transferred onto a PVDF membrane, stained with Brilliant Blue R-250, and sequenced. Five major V8 fragments were obtained, which were subjected to partial amino acid sequence analysis at their N-termini and these sequences presented in Table 1. Based upon these sequences, degenerate primers were used to probe a pea cDNA library to obtain cDNA fragments encoding these amino acid sequences.
|
Cloning of cDNA encoding HMP, and the complete primary structure of HMP
Initially, a cDNA corresponding to fragment V8-6 and downstream sequences was obtained, which encoded 700 amino acids from the C-terminus, and this was submitted (DDBJ/EMBL/GenBank AB055904). RACE-PCR was then performed to clone the 5' upstream region (nt 1–1065: DDBJ/EMBL/GenBank AB078602), which was contiguous with AB055904 (nt 1066–3379), and is shown in Fig. 4. The complete sequence encoding the partial sequence of V8-5 fragment was recovered at the junction (MRVLN) of these two clones. The total length of the combined cDNA was 3379 nt and it encoded a polypeptide of 989 amino acids, starting at nt 195 and ending at nt 3164 (DDBJ/EMBL/GenBank AB078603), with a molecular weight of over 108 kDa and a calculated pI of 7.13. All of the proteolytic fragments presented in Table 1 were found in this encoded polypeptide (underlined regions). The encoded polypeptide was composed of four SNc domains (shaded) preceding a Tudor domain (shaded region with bold letters), with an additional (partial) SNc domain (shaded) downstream from the Tudor domain. A search of the non-redundant database showed only one other identified plant protein, which was isolated from rice endosperm by Sami-Subbu et al, (2001). Additionally, the Arabidopsis genomic project putatively identified apparent homologues of human p100, and several animal proteins with similarity to HMP were found.
|
Since so little is known about these proteins, especially in plants, it was decided to do a comparative study of most of those that could be found and the results are summarized in Fig. 5, which shows the domain structures of 4SNc-Tudor proteins from pea (Psat) BAB32793; Arabidopsis (Athal) CAB87924; rice (Osat); Ajellomyces capsulatus (Acap) CAA06786; Aspergillus fumigatus (Afum), TIGR_5085; Candida albicans (Calb), SDSTC_5476; a fission yeast, Schizosaccharomyces pombe (Spom), CAB39904; mouse (Mmus) AAH07126; rat (Rnor) AAB41439; human (Hsap) AAA80488; chicken (Ggal) AAL27548; fly (Dmel) AAF47366; nematode (Cele) AAA81130. All of these proteins possessed 4 SNc domains in sequence followed by the Tudor domain (Fig. 5). It was noted that the chicken orthologue was incomplete at the 5' end, since no initiation codon was evident, thus the first SNc domain was absent. A fifth, SNc domain downstream (dSNc) from the Tudor domain was found in all proteins except the one from fission yeast, S. pombe. This does not seem to be the result of incomplete sequencing, since the genomic sequence contains the initiation codon, the apparent full ORF as well as some downstream 3' UTR and was described as complete by the authors (AL049498). Interestingly, another yeast (that forms pseudohyphae), Candida albicans did have this dSNc domain. In Tudor proteins from mammals (mouse, rat, human), this dSNc domain was also present, but severely truncated, while in plants (pea, Arabidopsis, rice) it was also somewhat truncated. Interestingly, in the Tudor proteins from widely diverse, non-mammalian animals (bird, fly, and nematode), this dSNc domain was full length, and it overlapped with the Tudor domain. It is apparent that these 4SNc-Tudor proteins have a common arrangement of four SNc domains followed by a Tudor domain, but they differ markedly in their dSNc domains, and the latter seems to be characteristic of the phylogenetic class of the source organism. Based on the dSNc domain, five classes can be described: those lacking the dSNc domain (fission yeast); those with a severely truncated version (mammals); those with a partially truncated version (plants), those with another partially truncated version (2 fungi and 1 yeast with pseudohyphae); and those with a full dSNc domain that overlaps the Tudor domain (non-mammalian animals).
|
Phylogenetic analysis of 4SNc-Tudor proteins between several classes of organisms
A more detailed phylogenetic analysis of these 4SNc-Tudor proteins is given in Fig. 6, where the amino acid identities and the similarities of proteins were compared: higher plants [pea, Arabidopsis (Athal1, 2), rice]; fungi [Ajellomyces capsulatus (Acap) Aspergillus fumigatus (Afum); Candida albicans (Calb); Schizosaccharomyces pombe (Spom)]; non-mammalian animals, a nematode (Cele), fly (Dmel), bird (Ggal)]; and mammals, rat (Rnor), mouse (Mmus1, 2), and human (Hsap1, 2)]. The results in Fig. 6 comparing the identities/similarities of the entire SNc-Tudor proteins, show that vertebrate proteins form a distinct, closely related group (lower right) while higher plants form another distinct, but less closely related group (upper left), with fungi and invertebrates comprising an intermediate group.
|
The phylogenetic tree of HMP proteins from these divergent classes of organisms was consistent with that expected from the evolutionary distance between these species (Fig. 7). Figure 7 also shows the interrelatedness of the five classes of 4SNc-Tudor proteins as described in Fig. 5: the mammalian (MM) type with a partial dSNc domain localized to the nucleus; the overlapping (OP) type in which the dSNc is full length and overlaps with the Tudor domain; the higher plant (HP) type with a partial dSNc and localized to the cyoskeleton; the fungal (FG) type with a partial dSNc, but distinct from both mammalian and higher plant types; and the fission yeast (FY) type which totally lacks a dSNc domain.
|
Detection and subcellular distribution of proteins antigenic to pea HMP in rice and wheat
In order to find out if other plants contained a similar Tudor protein with similar subcellular distribution to HMP of peas, an antibody was generated to the chromatographically pure protein (Fig. 3A, lane GF) in rat, as previously described (Shibata et al., 2002), and the data are shown in Fig. 8. An antigenic protein of slightly larger mass than in pea (1) is present in rice (2) and wheat (3). It is most abundant in the cytoskeleton fraction (lanes 1c, 2c, 3c) less so in the nuclei (lanes 1b, 2b, 3b) and least in the supernatant fraction (lanes 1d, 2d, 3d). A single streptavidin-binding protein is present in rice (2) and in wheat (3), while two such proteins are present in pea (1).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
One of the proteins isolated from the pea cytoskeleton fraction was a high molecular weight protein (HMP) of 110 kDa (Abe et al., 1992), which co-localized with the ribosome–cytoskeleton complex isolated by centrifugation (Ito et al., 1994). This protein (HMP) was isolated by heparin affinity column chromatography and shown to bind heparin quite tightly, since it did not elute from the column until about 390 mM KOAc (Shibata et al., 1999). Its partial primary amino acid sequence was reported and, for the first time, the existence in plant cells of a Tudor protein with multiple SNc domains (Accession No. AB055904) was shown, while a somewhat similar protein has recently been isolated from the cytoskeleton fraction from rice endosperm (Sami-Subbu et al., 2001). The Arabidopsis genomic project has identified another putative Tudor protein (AL163912) based on the similarity of its domain structure to that of a human Tudor protein (p100).
In the present study, it has been shown that this Tudor domain protein from peas is abundant in the cytoskeleton fraction, but essentially absent from the nucleus (Fig. 1A). This strongly implies that it is not a nuclear-localized transcription factor, as it appears to be in human tissue (Tong et al., 1995), but is more likely to be involved in cytoskeleton-related events, including mRNA transport or translation (Davies et al., 2001). The rice endosperm protein was also localized to the cytoskeleton and it was thought to have a role in RNA transport (Sami-Subbu et al., 2001). In this context, it is shown that HMP elutes from a heparin affinity column in a major peak (Fig. 1B) that absorbs at both 280 nm (protein) and 260 nm (nucleic acid) and co-elutes with ribosomes if they had not previously been removed by centrifugation (data not shown), thus implying that HMP may also bind to ribosomes, rRNA and/or ribosomal proteins. It was also found that ribosomal proteins were present in the void fraction of a superose column (Fig. 3B) and they co-eluted with HMP. It was previously reported that this protein co-localized with a ribosome–cytoskeleton–membrane complex separated on sucrose gradients (Ito et al., 1994), which also suggests that HMP has affinity to ribosomes, at least in the presence of the low ionic strength buffer, CSB. However, under high ionic strength conditions, HMP dissociates from both ribosomes and the cytoskeleton (when separated by gel filtration column chromatography in the presence of at least 400 mM KOAc as seen in Fig. 2B, C). After superose column chromatograpy, HMP elutes as a 220 kDa complex (Fig. 2B), implying that it forms a dimer in vitro, and perhaps in vivo. Isoelectric focusing gels show that there are four major isotypes of HMP with pIs of 6.04, 6.13, 6.24, and 6.34 (Fig. 3C), with an almost constant pH difference of 0.09–0.11 between each isotype (Fig. 3). This pattern of isotype may result from increasing extents of phosphorylation, as found with the elongation factor binding protein (eIF4E-BP1) in mammalian cells (Duncan and Song, 1999).
The complete structure of HMP from peas was determined (Fig. 4), and it was confirmed that it has four SNc domains followed by a Tudor domain, which was called a 4SNc-Tudor protein, although it does also have a fifth, incomplete downstream (dSNc) domain. The purified protein yielded several fragments after hydrolysis by V8 protease or CNBr (Table 1), and these fragments were all consistent with the parent protein being a 4SNc-Tudor domain protein (Fig. 4). The Arabidopsis genomic project identified two genes encoding two similar 4SNc-Tudor proteins and, based on comparisons with animal proteins, tentatively identified these gene products (CAB87924, BAB10078) as transcription factors localized in the nucleus. However, experimental evidence is totally lacking for this assignment (Sato et al., 1998). Sami-Subbu et al. (2001) reported that the rice 4SNc-Tudor protein is present in the cytoskeleton, but not in the nucleus of developing rice endosperm. When SNc domain proteins are present in nuclei, their most likely function is to bind single-stranded DNA or newly formed mRNA during transcription (Tong et al., 1995). The human 4SNc-Tudor protein (p100) is localized primarily to the nucleus and was found to bind to acidic proteins, probably to provide protein–protein interactions for its function as a transcription co-factor (Tong et al., 1995), whereas the plant homologue is cytoskeleton-bound, but may bind RNA (Sami-Subbu et al., 2001). Recently, however, the 100 kDa 4SNc-Tudor protein in human has also been found in the endoplasmic reticulum (Broadhurst and Wheeler, 2001), perhaps in association with polysomes on the RER. Alternatively, the 4SNc-Tudor protein might be involved in the shuttling of RNA (Schroder et al., 1986, 1988) and/or ribosomes between the nucleus and the cytoplasm both in human and plant cells. In addition, a homologue to human EBNA2-activated transcription cofactor P100 has also been identified in a pathogenic fungus, but the location of the encoded protein has not been experimentally determined (Porta et al., 1999).
This 4SNc-Tudor protein has homologues in a diverse array of organisms including fungi, lower animals, vertebrates, and higher plants (Fig. 5). The molecular mass of this class of 4SNc-Tudor proteins is about 100 kDa in humans and 110 kDa in plants, and the number of amino acids ranges from a high of 986–1050 (higher plants), 914–926 (lower animals), 885–910 (mammals), 878–890 (fungi) to 714 (birds). As expected, these results (Figs 5, 6) show that 4SNc-Tudor proteins from pea, rice and Arabidopsis are very similar while homology between plant and animal 4SNc-Tudor proteins is far less. In human and mouse there were two almost identical sequences for 4SNc-Tudor proteins, but these appeared to be alleles (7q31.3), and only in Arabidopsis do there appear to be two totally distinct genes.
Despite the highly conserved overall domain structure (Fig. 5), considerable variation exists between individual domains, especially the fifth (downstream) SNc domain. In plants, this domain is incomplete and appears to form a coiled-coil structure, which enables it to bind to the cytoskeleton (Sami-Subbu et al., 2001), whereas the human orthologue lacks this structure and the protein is localized to the nucleus (Tong et al., 1995).
The downstream (dSNc) domain is totally absent (yeast), severely degenerate (mammals), partly truncated (higher plants) or complete but overlapping the Tudor domain (chicken, fly, nematode). The dSNc domain in HMP is absent in a fission yeast (Moreno et al., 1991), Schizosaccharomyces pombe (Fig. 5), but in budding yeast, Saccharomyces cerevisiae databases indicate HMP is totally absent, while another yeast (Candida albicans) that can form pseudohyphae (McGeady et al., 2002), had a 4SNc-Tudor similar to the fungal type. Together, these data suggest that HMP may play a crucial role in ‘normal’ cell division, fission, and elongation, which is a cytoskeleton-dependent process, and not in cell budding. The phylogenetic tree (Fig. 7), which does not emphasize the dSNc domain, shows that ancestral yeasts possessed the common (fungal) type of 4SNc-Tudor protein, while higher plants separated earlier from the remaining organisms than separation of animal, fungi, and yeasts. It is anticipated that a more detailed phylogeny of 4SNc-Tudor proteins may provide crucial insight into the evolution of fungi and yeasts. Finally, given the potential importance of the dSNc domain in determining intracellular localization, it is hypothesized that sequence divergence in this motif has profound functional significance. Future studies will elucidate the role of this region on intracellular localization and function of 4 SNc-Tudor proteins.
Experimental evidence for peas, rice, and wheat (Fig. 8) indicates the primary location of 4SNc-Tudor protein is the cytoskeleton, but it is also present in nuclei. A protein antigenic to the 4SNc-Tudor protein in dark-grown wheat and rice seedlings was also detected (Fig. 8). However, previous workers using an antibody raised against SDS–PAGE-purified protein (Sami-Subbu et al., 2001) reported that the 4SNc-Tudor protein is present only in the endosperm of rice. A possible reason for why they failed to detect 4SNc-Tudor in other tissues is because they used SDS–PAGE-purified proteins from rice to generate antibodies. By contrast, the antibody made against liquid chromatography purified protein gave a high titre (1:6000 for 0.1 ng protein), and was strong enough to detect the protein in other plants. However, the much darker bands from pea (Fig. 8, panel 1) might not result from higher amounts of the corresponding antigen in that tissue, but of higher activity of the antibody to the protein from pea.
|
Acknowledgements |
|---|
This research was supported financially by a Sasagawa Scientific Research Grant from the Japan Science Society (to KS) and by the NC Agricultural Research Service (no. 06446 to ED). Sequence data of Aspergillus fumigatus (TIGR_5085) was obtained from The Institute for Genomic Research (TIGR) website at http://www.tigr.org. Sequencing of Aspergillus fumigatus was accomplished with support from NIH/NIAID/Wellcome Trust. We also thank Todd Grey of the NewYork Department of Health and Ron Sederoff from NCSU for their critical review of this manuscript.
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Abe S, Davies E. 1991. Isolation of F-actin from pea stems: evidence from fluorescence microscopy. Protoplasma 163, 51–61.[CrossRef][Web of Science]
Abe S, Davies E. 1995. Methods for isolation and analysis of the cytoskeleton. In: Galbraith DW, Bourque DP, Bohnert HJ, eds. Methods in plant cell biology, Part B. California: Academic Press, 223–236.
Abe S, Ito Y, Davies E. 1992. Co-sedimentation of actin, tubulin and membranes in the cytoskeleton fractions from peas and mouse 3T3 cells. Journal of Experimental Botany 43, 941–949.
Abe S, Ito Y, Davies E. 1994. Association of cytoskeletal proteins in the membrane-bound polysome fraction from peas using conventional polysome isolation buffers. Plant Physiology and Biochemistry 32, 547–554.[Web of Science]
Abe S, Ito Y, Davies E. 1995. Isolation of a heparin-sensitive, ribosome-sedimenting factor from the cytoskeleton fractions of peas and corn. Plant Physiology and Biochemistry 33, 463–470.[Web of Science]
Bradford MM. 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein–dye binding. Analytical Biochemistry 72, 248–254.[CrossRef][Web of Science][Medline]
Broadhurst MK, Wheeler TT. 2001. The p100 coactivator is present in the nuclei of mammary epithelial cells and its abundance is increased in response to prolactin in culture and in mammary tissue during lactation. Journal of Endocrinology 171, 329–337.[Abstract]
Cleveland DW, Fischer SG, Krischner MW, Laemmli UK. 1977. Peptide mapping by limited proteolysis in sodium dodecyl sulphate and analysis by gel electrophoresis. Journal of Biological Chemistry 252, 1102–1106.
Davies E, Abe S, Larkins BA, Clore AM, Quatrano RS, Weidner S. 1998. The role of the cytoskeleton in plant protein synthesis. In: Bailey-Serres J, Gallie DR, eds. a look beyond transcription: mechanisms determining mRNA stability and translation in plants. American Society of Plant Physiologists, 115–124.
Davies E, Fillingham BD, Abe S. 1996. The plant cytoskeleton. In: Hesketh JE, Pryme IF, eds. The cytoskeleton, Vol. 3. USA: JAI Press, 405–449.
Davies E, Stankovic B, Azama K, Shibata K, Abe S. 2001. Novel components of the plant cytoskeleton: a beginning to plant ‘cytomics’. Plant Science 160, 185–196.
Duncan RF, Song HJP. 1999. Striking multiplicity of eIF4E-BP1 phosphorylated isoforms identified by 2D gel electrophoresis—regulation by heat shock. European Journal of Biochemistry 265, 728–743.[Web of Science][Medline]
Ito Y, Abe S, Davies E. 1994. Co-localization of cytoskeleton proteins and polysomes with a membrane fraction from peas. Journal of Experimental Botany 45, 253–359.
Jahnen-Dechent W, Simpson RJ. 1990. A method for preparing proteins and peptides for microsequencing. Plant Molecular Biology Reporter 8, 92–103.
Keefe LJ, Quirk S, Gittis A, Sondek J, Lattman EE. 1994. Accommodation of insertion mutations on the surface and in the interior of staphylococcal nuclease. Protein Science 3, 391–401.[Web of Science][Medline]
Keefe LJ, Sondek J, Shortle D, Lattman EE. 1993. The alpha aneurism: a structural motif revealed in an insertion mutant of staphylococcal nuclease. Proceedings of the National Academy of Sciences, USA 90, 3275–3279.
McGeady P, Logan DA, Wansley DL. 2002. A protein-farnesyl transferase inhibitor interferes with the serum-induced conversion of Candida albicans from a cellular yeast form to a filamentous form. FEMS Microbiology Letters 213, 41–44.[CrossRef][Web of Science][Medline]
Moreno S, Klar A, Nurse P. 1991. Molecular genetic analysis of fission yeast Schizosaccharomyces pombe. Methods in enzymology 194, 795–823.[Web of Science][Medline]
Ponting CP. 1997. Tudor domains in proteins that interact with RNA. Trends in Biochemical Science 2, 51–52.
Porta A, Colonna-Romano S, Callebaut I, Franco A, Marzullo L, Kobayashi GS, Maresca B. 1999. A homologue of the human 100 kDa protein (p100) is differentially expressed by Histoplasma capsulatum during infection of murine macrophages. Biochemical and Biophysical Research Communications 254, 605–613.[CrossRef][Web of Science][Medline]
Sato S, Kaneko T, Kotani H, Nakamura Y, Asamizu E, Miyajima N, Tabata S. 1998. Structural analysis of Arabidopsis thaliana chromosome 5. IV. Sequence features of the regions of 1,456,315 bp covered by 19 physically assigned P1 and TAC clones. DNA Research 5, 41–54.[Abstract]
Schroder HC, Diehl-Seifert B, Rottmann M, Messer R, Bryson BA, Agutter PS, Muller WE. 1988. Functional dissection of nuclear envelope mRNA translocation system: effect of phorbol ester and a monoclonal antibody recognizing cytoskeletal structures. Archives of Biochemistry and Biophysics 261, 394–404.[CrossRef][Web of Science][Medline]
Schroder HC, Rottmann M, Bachmann M, Muller WE, McDonald AR, Agutter PS. 1986. Proteins from rat liver cytosol which stimulates mRNA transport: purification and interactions with the nuclear envelope mRNA translation system. European Journal of Biochemistry 59, 51–59.[Web of Science][Medline]
Sami-Subbu R, Choi S-B, Wu Y, Wang C, Okita TW. 2001. Identification of a cytoskeleton-associated 120 kDa RNA-binding protein in developing rice seeds. Plant Molecular Biology 46, 79–88.[CrossRef][Web of Science][Medline]
Shibata K, Abe S, Davies E. 2001. Structure of the coding region and mRNA variants of the apyrase gene from pea (Pisum sativum). Acta Physiologiae Plantarum 23, 3–13.[Web of Science][Medline]
Shibata K, Abe S, Yoneda M, Davies E. 2002. The subcellular distribution and isotypes of a 49 kDa apyrase from pea (Pisum sativum L. var. Alaska). Plant Physiology and Biochemistry 40, 407–415.[CrossRef][Web of Science]
Shibata K, Morita Y, Abe S, Stankovic B, Davies E. 1999. Apyrase from pea stems: isolation, purification, characterization, and identification of a NTPase from the cytoskeleton fraction of pea stem tissue. Plant Physiology and Biochemistry 37, 881–888.[CrossRef][Web of Science]
Tong X, Drapkin R, Yalamanchili R, Mosialos G, Kieff E. 1995. The Epstein–Barr virus nuclear protein 2 acidic domain forms a complex with a novel cellular coactivator that can interact with TFIIE. Molecular and Cellular Biology 15, 4735–4744.[Abstract]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J. Yang, T. Valineva, J. Hong, T. Bu, Z. Yao, O. N. Jensen, M. J. Frilander, and O. Silvennoinen Transcriptional co-activator protein p100 interacts with snRNP proteins and facilitates the assembly of the spliceosome Nucleic Acids Res., July 26, 2007; 35(13): 4485 - 4494. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. D. X. Chuong, A. G. Good, G. J. Taylor, M. C. Freeman, G. B. G. Moorhead, and D. G. Muench Large-scale Identification of Tubulin-binding Proteins Provides Insight on Subcellular Trafficking, Metabolic Channeling, and Signaling in Plant Cells Mol. Cell. Proteomics, October 1, 2004; 3(10): 970 - 983. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||