HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 281, Issue 42, 31457-31466, October 20, 2006

This Article Abstract Full Text (PDF) All Versions of this Article: 281/42/31457    most recent M606643200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Liu, Y. Articles by Chang, A. Search for Related Content PubMed PubMed Citation Articles by Liu, Y. Articles by Chang, A. Social Bookmarking                      What's this?

Multiple Degradation Pathways for Misfolded Mutants of the Yeast Plasma Membrane ATPase, PMA1^*

From the Department of Molecular, Cellular & Developmental Biology, University of Michigan, Ann Arbor, Michigan 48109

Received for publication, July 13, 2006 , and in revised form, August 15, 2006.

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
To understand protein sorting and quality control in the secretory pathway, we have analyzed intracellular trafficking of the yeast plasma membrane ATPase, Pma1. Pma1 is ideal for such studies because it is a very abundant polytopic membrane protein, and its localization and activity at the plasma membrane are essential for cell viability and growth. We have tested whether the cytoplasmic amino- and carboxyl-terminal domains of Pma1 carry sorting information. As the sole copy of Pma1, mutants truncated at either NH₂ or COOH termini are targeted at least partially to the plasma membrane and have catalytic activity to sustain cell viability. The mutants are also delivered to degradative pathways. Strikingly, NH₂- and COOH-terminal Pma1 mutants are differentially recognized for degradation at distinct cellular locales. COOH-terminal mutants are recognized for destruction by endoplasmic reticulum-associated degradation. By contrast, NH₂-terminal mutants escape detection by endoplasmic reticulum-associated degradation entirely, and undergo endocytosis for vacuolar degradation after apparently normal cell surface targeting. Both NH₂- and COOH-terminal mutants are conformationally abnormal, as revealed by increased sensitivity to tryptic cleavage, but are able to assemble to form oligomers. We propose that different quality control mechanisms may assess discrete domains of Pma1 rather than a global conformational state.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
There are multiple mechanisms for recognition and destruction of misfolded, and/or unassembled proteins in the secretory pathway (1). Endoplasmic reticulum (ER)²-associated degradation (ERAD) is one major pathway that uses conformational recognition by chaperones to select defective proteins for ubiquitination via ER-associated ubiquitin ligases; these proteins then undergo dislocation from the endoplasmic reticulum, and destruction by the 26 S proteasome (2). In addition to ERAD, however, Golgi-based targeting of proteins for lysosomal/vacuolar destruction has been described. Moreover, there is accruing evidence for plasma membrane quality control in which selected proteins are identified for turnover from the cell surface via endocytosis and lysosomal/vacuolar delivery (3). The molecular mechanisms by which proteins of the secretory pathway are recognized by each of these degradative pathways is of considerable interest.

We have studied the yeast plasma membrane ATPase, Pma1, as a model protein to understand protein sorting and quality control in the secretory pathway. Pma1 is a H⁺-ATPase that pumps protons out of the cell to generate the membrane potential and regulate cytosolic pH (4). Therefore, proper delivery of Pma1 to the cell surface is essential for cell viability. Pma1 belongs to the P-type ATPase family of ion transporters that includes the Na⁺K⁺- and Ca²⁺-ATPases of mammalian cells (5, 6). The 8- and 2.6-Å structures of Neurospora Pma1 and mammalian sarcoplasmic reticulum Ca²⁺-ATPase have been determined (6, 7), and based on these data, a structural model for fungal Pma1 has been proposed (8). Like other P-type ATPases, Pma1 is embedded in the membrane by 10 transmembrane segments, and there are 3 major cytoplasmic domains that contain the nucleotide-binding and catalytic phosphorylation sites and have critical roles in catalytic activity. The amino and carboxyl termini are cytoplasmic; in several family members, these domains represent regulatory domains (9). For instance, the fungal H⁺ pumps have COOH-terminal regulatory domains that modulate activity by kinase-mediated phosphorylation (10). Pma1 forms a hexamer (7), or two hexamers may come together to form a dodecamer (11), and the COOH-terminal domain has also been proposed to participate in the oligomeric structure of Pma1 (8).

Newly synthesized wild-type Pma1 undergoes efficient intracellular transport to the plasma membrane where it acquires remarkable longevity (12). ERAD destroys misfolded Pma1 mutants; however, some Pma1 mutants have been characterized that escape ER retention and ERAD (1). For example, Pma1–10 is misfolded, and yet properly targeted to the plasma membrane; however, its turnover from the plasma membrane is increased by comparison with wild-type Pma1 (13). Of relevance to the question of ER export versus retention is the identification of ER export signals in the cytoplasmic COOH-terminal domain of some polytopic membrane proteins (14). Diacidic and dihydrophobic motifs appear to promote entry into COPII-coated vesicles, in some cases, acting combinatorially (14, 15). Moreover, arginine-based sorting motifs have been identified in the cytosolic domains of multimeric membrane proteins that become masked upon proper protein assembly, leading to ER exit of the complexes (16). To ask whether the cytoplasmic NH₂- and COOH-terminal domains of Pma1 carry sorting motifs, we constructed mutants truncated at either termini and analyzed their intracellular transport and stability at the cell surface.

In this paper, we report that mutants at either NH₂ and COOH termini are conformationally abnormal, as revealed by increased sensitivity to tryptic cleavage. Nevertheless, these Pma1 mutants are differentially recognized for degradation at distinct cellular locales. NH₂-terminal mutants escape ERAD entirely; after apparently normal targeting to the cell surface, they are unstable at the cell surface and undergo endocytosis for vacuolar degradation. By contrast, COOH-terminal mutants are recognized by ERAD, although some of the protein is able to escape to the plasma membrane and sustain cell viability. Although trafficking through the secretory pathway is clearly affected by COOH-terminal truncation, Pma1 oligomerization is not impaired. We propose that different quality control mechanisms may assess discrete domains of Pma1 rather than a global conformational state.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Media and Strains—Standard yeast media and genetic manipulations were as described (17). Yeast transformations were performed by the lithium acetate method (18). Strains are W303-derived (ura3-1 leu2-3,112 his3-11 trp1-1 ade2-1 can1-100; also referred to as F1105) unless otherwise noted. Strains with COOH- and NH₂-terminal truncations in PMA1 were generated by PCR-based gene modification (19).

SSX5 (40C diploid) is a heterozygote with one wild-type PMA1 and one chromosomal copy of a truncation of the carboxyl-terminal 40 residues of PMA1 followed by a triple HA tag; to generate the strain, a W303 diploid was transformed with products made by PCR amplication of pFA6a-3HA-HIS3MX6 (19) using primers SS11 and SS13. All primer sequences are available upon request. SSX5-1a is a MATa ura3-1 leu2-3,112 his3-11 trp1-1 ade2-1 can1-100 pma140C-HA::HIS3 haploid generated by sporulation and tetrad dissection of SSX5. SSX9 (30C diploid) is a heterozygote with wild-type PMA1 and a truncation of the carboxyl-terminal 30 residues of PMA1 followed by a triple HA tag; this strain was generated as described for 40C except that primers SS11 and SS19 were used to generate PCR products for transformation. SSX9-1b is a MATa ura3-1 leu2-3,112 his3-11 trp1-1 ade2-1 can1-100 pma130C-HA::HIS3 haploid generated by sporulation and tetrad dissection of SSX9. pma130C-HA::HIS3 fails to grow at 37 °C; this phenotype is complemented by PMA1. SSY3 was similarly constructed except that primers SS11 and SS10 were used to generate PCR products for transformation of haploid W303 (F1105). Correct integration was confirmed by checking insertion of the HA tag by Western blot and/or PCR. SSY38 (MATa ura3-1 leu2-3,112 his3-11 trp1-1 ade2-1 can1-100 pma130C-HA::HIS3 end4::clonNAT) was made by transformation of SSX9-1b with PCR products to knock-out END4 marked by resistance to clonNAT; primers 294 and 295 were used to amplify pAG25. SSY39 (MATa ura3-1 leu2-3,112 his3-1 trp1-1 ade2-1 can1-100 pma130C-HA::HIS3 pep4::LEU2) was made by transformation of SSX9-1b with a PEP4 disruption construct pSN273 cut with SacI and XhoI (20). SSY24 was constructed by first transforming MHY1703 (MATa his3200 leu2-3,11 ura3-52 lys2-801 trp1-1 doa10::HIS3 hrd1::LEU2) (21) with pHT6 to swap the HIS3 marker for a TRP1 marker to generate SSY23 (22); SSY23 was then crossed with SSX5-1a, the diploid was sporulated and dissected to produce a MATa strain that is also pma140C::HIS3 hrd1::LEU2 doa10::TPR1. SSY25 was made by first transforming MHY552 (MATa his3200 leu2-3,112 ura3-52 lys2-801 trp1-1 ubc61::HIS3 ubc7::LEU2) with pHT6 to replace the HIS3 marker with TRP1, generating SSY15; SSY15 was then crossed with SSX5-1a, the diploid was sporulated and dissected to produce a MATa strain that is also pma140C::HIS3 ubc6::TRP1 ubc7::LEU2. Similarly, SSY37 (MATa ura3-1 leu2-3,112 his3-11 trp1-1 ade2-1 can1-100 pma130C-HA::HIS3 ubc6::TRP1 ubc7::LEU2) was made by crossing SSX9-2b (MATa) with SSY15. ACX134 is a W303 diploid transformed with PCR products to make a PMA1 knock-out marked by resistance to clonNAT; primers 397 and 398 were used to amplify the template pAG25 bearing natMX4 (obtained from Charlie Boone, University of Toronto, Toronto, Canada), conferring resistance to the antibiotic nourseothricin (clonNAT, Werner BioAgents) (23). The knock-out was confirmed by PCR. SSY29 is a MATa pma1::clonNAT^R (pXG29) haploid generated upon transformation of ACX134 followed by sporulation and tetrad dissection.

KKY39 is a heterozygous diploid with wild-type PMA1 and another chromosomal copy of an HA-tagged pma1 mutant with an NH₂-terminal truncation of 40 residues; the strain was generated by transformation of a W303 diploid with products made by PCR amplication of pFA6a-HIS3MX6-pGAL1–3HA (19) using primers 305 and 272. KKY44 is a HIS3⁺ haploid 40N strain generated by sporulation and dissection of KKY39. KKY40 is a heterozygous diploid with GAL1-HA-60N-pma1, generated as described for 40N except that primers 306 and 272 were used for PCR amplification. KKY37 is a heterozygous diploid with wild-type PMA1 and GAL1-HA-PMA1, constructed as described for 40N except that primers 303 and 272 were used for PCR amplification. KKX24-1B is a HIS3⁺ haploid HA-PMA1 strain generated by sporulation and dissection of KKY37. KKY93 is pma1::clonNAT^R (pKK107); KKY94 is pma1::clonNAT^R (pKK109). Both strains were constructed by plasmid shuffle: SSY29 was transformed with pKK107, a LEU2-marked centromeric plasmid bearing HA-tagged wild-type PMA1 or pKK109, a LEU2-marked centromeric plasmid bearing HA-tagged 4D2E/A mutant; loss of wild-type PMA1 (pXG29) was selected on 5-fluoroorotic acid (24). Other yeast strains used to examine the behavior of plasmid-borne pma1-4D2E/A are L3852 (MATa his3200 lys2201 leu2-3,112 ura3-52 ade2) and isogenic ACY67 (MATa his3200 lys2201 leu2-3,112 ura3-52 ade2 pep4) and RH266-1D (MATa end3-1 leu2 his4 ura3 bar1-1) (25).

Molecular Biology—pXG29 is a URA3-marked centromeric plasmid constructed by placing a 5-kb HindII-HindIII fragment containing PMA1 promoter and coding sequence from pAC4 into pRS316 (26). pKK107 and pKK109 are LEU2-marked centromeric plasmids bearing HA-tagged PMA1 and HA-tagged pma1-4D2E/A, respectively, under control of the native promoter. pKK107 is derived from a URA3-marked YIp bearing a 4.5-kb SacI-XhoI fragment containing MET3-HA-PMA1 (pKK98). A 0.5-kb SacI-XmaI fragment was excised to replace the MET3 promoter with an 0.8-kb fragment containing a PMA1 promoter sequence amplified with introduced sites using primers 395 and 396. pKK103, a URA3-marked YIp bearing pma1-4D2E/A under the control of MET3, constructed by using QuikChange site-directed mutagenesis (Stratagene, La Jolla, CA) was used to introduce D39A/D40A/D41A with primers 393 and 394 followed by D42A/E48A/E49A changes with primers 399 and 400. PCR mutagenesis was performed using as template pKK74, an excised 1.8-kb fragment bearing MET3-HA-PMA1 up to the unique BamHI site, derived from pCC2 (27). Mutations were confirmed by DNA sequencing. pKK98 and pKK103 were transformed into yeast after cutting with NcoI to linearize the plasmids at the URA3 marker. Construction of pKK109 was similar to that of pKK107 except that the MET3 promoter was excised as a 0.5-kb SacI-XmaI fragment for replacement with PMA1 promoter sequence.

For co-immunoprecipitation experiments, inducible constructs were used. pND542 is a LEU2-marked centromeric plasmid bearing HA-tagged wild-type Pma1 under the control of the GAL1 promoter, described in Ref. 28. pWQ3 is a URA3-marked centromeric plasmid bearing myc-tagged wild-type Pma1 under GAL1 control. pSS16 is a URA3-marked centromeric plasmid bearing GAL1-myc-pma140C; to construct this, pWQ3 was used as a PCR template to introduce a stop codon TAA after Thr⁸⁷⁸ using primers SS31 and SS32. pSS17 is a LEU2-marked centromeric plasmid bearing GAL1-HA-pma140C; this was constructed by using pND542 as template for PCR to introduce a stop codon after Thr⁸⁷⁸ using primers SS31 and SS32.

Indirect Immunofluorescence—Indirect immunofluorescence was as described previously (29). Cells were fixed with 4% formaldehyde for 2 h at room temperature. Cells were spheroplasted with zymolyase 100T (ICN) and permeabilized with methanol and acetone. Cells were stained with monoclonal anti-HA (Covance, Inc.) or monoclonal anti-Pma1 (gift from John Aris, University of Florida), followed by fluorescent CY3, Texas Red, and/or dichlorotriazinylaminofluorescein-conjugated secondary antibodies (Jackson Immunochemicals). Cells were visualized with an Olympus fluorescence microscope and images were collected with a Hamamatsu Orca CCD camera. Anti-Kar2 antibody was a gift from Mark Rose (Princeton University, Princeton, NJ).

Protein Induction, Metabolic Labeling, Cell Fractionation, Trypsinolysis, and Western Blot—For induction of GAL1-regulated constructs, cells were grown overnight in medium containing 2% raffinose. Mid-log cultures were then resuspended in medium containing 2% galactose for 2–4 h. For MET3-regulated constructs, cells were grown overnight in medium with 600 µM methionine to repress MET3; exponentially growing cells were then washed with water and transferred to methionine-free medium to induce protein synthesis.

For metabolic labeling, cells were grown overnight without cysteine and methionine (usually minimal medium) to mid-log phase. Cells were resuspended in fresh medium at a density of 1 A₆₀₀/ml and incubated at room temperature for 15 min before pulse labeling with Expre³⁵S³⁵S (0.4 mCi/A₆₀₀ cells) (PerkinElmer Life Sciences). Cells were labeled for 10 min before chasing with an equal volume of synthetic complete medium supplemented with 20 mM methionine and cysteine. At various times of chase, aliquots were removed and added to 10 mM azide on ice. Lysate was prepared by vortexing with glass beads (30), and immunoprecipitation was normalized to acidprecipitable counts/min. Immunoprecipitations with anti-HA or anti-Pma1 antibodies were analyzed by SDS-PAGE and fluorography.

Cell fractionation by Renografin density gradients were as described (31). RenoCal-76 was substituted for Renografin-76. Fourteen fractions were collected and diluted with buffer (50 mM Tris, pH 7.5, 1 mM EDTA), membranes were pelleted by centrifugation at 100,000 x g for 1 h, and analyzed by Western blotting. Markers for ER, Golgi, and plasma membrane were localized using antibodies against Sec61 (from Randy Schekman, University of California, Berkeley, CA), Gda1 (from Greg Payne, UCLA, Los Angeles, CA), and Gas1 (from Howard Riezman, University of Geneva, Geneva, Switzerland). Quantitation was performed using NIH Image on scanned Western blots.

Limited trypsinolysis was performed as described (32). Total membranes were generated by centrifugation of cell lysate at 100,000 x g for 1 h. Membranes were resuspended in buffer (20 mM HEPES, pH 7.5, 100 mM KCl, 2 mM dithiothreitol), and incubated at 30 °C at a 1:25 trypsin:protein ratio. At various times, 10% ice-cold trichloroacetic acid was added, protein was precipitated on ice for >20 min, and precipitates were washed with cold acetone. Samples were analyzed by Western blot with anti-Pma1 and/or anti-HA antibodies. For Western blot of lysate, samples were normalized to protein content measured by the Bradford assay. Immune complexes were visualized by an ECL detection system.

Gel Filtration Chromatography—Lysate (1.5 mg of protein) was solubilized with 1% digitonin in 50 mM Tris, pH 7.5, 150 mM NaCl, 5 mM EDTA. Insoluble material was pelleted by centrifugation at 100,000 x g for 1 h. One ml of extract was fractionated over a gel filtration column (Sephacryl S300HR) with a range of 10⁴ to 1.5 x 10⁶ daltons using the FPLC system (Amersham Biosciences). A void volume of 39 ml was estimated using blue dextran. Fractions (40 x 1 ml) were collected from 36 ml. Protein was precipitated with trichloroacetic acid, and every two fractions were pooled by analysis by Western blot. For determination of molecular weight, a high molecular weight calibration kit (Amersham Biosciences) was fractionated.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Pma1 Truncation Mutants—Strains were constructed in which chromosomal PMA1 was truncated at the COOH terminus. According to hydropathy predictions, removal of the last 40 amino acid residues, 40C, should result in loss of the entire COOH-terminal cytoplasmic domain; removal of 30 residues, 30C, was also constructed to avoid possible perturbation of the preceding transmembrane segment. Heterozygous diploid strains were generated by PCR-mediated transformation to generate one chromosomal copy of a PMA1 truncation that was also HA-tagged at the COOH terminus and marked by HIS3. Sporulation and tetrad dissection resulted in viable haploid 30C and 40C pma1 mutants. Haploid 30C and 40C strains were also readily generated by plasmid shuffle (not shown).

In parallel with truncations at the COOH terminus, a series of truncations were also made at the NH₂ terminus. These mutants made by PCR-mediated transformation of diploid cells are HA-tagged at the NH₂ terminus, marked with HIS3, and under the control of the GAL1 promoter. Diploids were sporulated and dissected onto galactose-containing medium to determine the effect of NH₂-terminal truncations on cell growth. Like the COOH-terminal truncations, strains without 40 residues at the NH₂ terminus, 40N, are viable with no obvious growth defect (not shown); however, there is impaired trafficking (see below). 60N mutants are not able to support growth as the sole copy of Pma1 (not shown), in agreement with a previous report (33).

Localization of Pma1 COOH-terminal Truncation Mutants—We expected Pma1 COOH-terminal truncation mutants to localize at the plasma membrane because haploid cells with these mutations are viable. Indirect immunofluorescence localization of HA-tagged 40C was therefore surprising because significant perinuclear staining was seen with anti-HA antibody, overlapping with the ER marker Kar2 (Fig. 1A). By contrast, anti-Pma1 antibody revealed predominant plasma membrane staining (Fig. 1A) with faint perinuclear staining in most of these cells. 30 Pma1 localization was observed at the cell surface as well as in punctate spots; the cell surface and punctate staining pattern is similar with both anti-HA and anti-Pma1 antibodies (Fig. 1A, left panels). In a pep4 background in which vacuolar proteolysis is prevented (34), 30C Pma1 staining was observed in vacuoles, seen as indentations by differential interference contrast, whereas vacuolar localization was not seen for 40C Pma1 (not shown). By contrast with the mutant strains, wild-type Pma1 HA-tagged at the NH₂ terminus appeared exclusively localized at the plasma membrane (Fig. 2A); wild-type Pma1 HA-tagged at the COOH terminus appeared predominantly cell surface-localized (Fig. 1A, top right panel).

To quantitate cell surface and intracellular 30C and 40C Pma1, cell lysate was fractionated on Renografin density gradients. As shown previously, Renografin density gradients effectively separate plasma membrane from intracellular membranes (31). The distribution of membrane markers analyzed by Western blot is in Fig. 1B, showing that the plasma membrane marker Gas1 (triangle) fractionates in denser gradient fractions 9–12, whereas intracellular membrane markers, both Golgi GDPase (open circle) and ER Sec61 (diamond), are in fractions 2–5 at the top of the gradient. Peak fractions containing HA-tagged wild-type Pma1 (square) are coincident with the Gas1 marker (Fig. 1B). Similarly, the preponderance of Pma130C is present in plasma membrane-enriched fractions but there is also a fraction coincident with intracellular membranes (Fig. 1B, right panel, square), consistent with the puncta seen by indirect immunofluorescence (Fig. 1A). Fractionation of 40C Pma1 and detection by anti-HA antibody results in a majority of 40C Pma1 (square) distributed in intracellular membrane fractions 2–5, coincident with the ER marker Sec61. Detection of 40C Pma1 with anti-Pma1 monoclonal antibody (closed circle) revealed the converse distribution with the majority in fractions 9–12 with a small shoulder trailing to the top of the gradient, in agreement with the indirect immunofluorescence results in Fig. 1A. These data are consistent with the idea that the HA epitope at the COOH terminus of 40C Pma1 is removed after ER export, and the majority of 40C Pma1 is detected at the plasma membrane by the monoclonal antibody.

Localization of NH₂-terminal Truncation Mutants—Localization of newly synthesized 40N was detected by indirect immunofluorescence by inducing synthesis of HA-tagged 40N Pma1 in a heterozygous diploid. 40N staining is at the cell surface as well as at the vacuole, seen as indentations by differential interference contrast microscopy, whereas newly synthesized wild-type Pma1 is exclusively at the plasma membrane (Fig. 2A). 60N staining appeared over vacuoles and as intracellular dots surrounding vacuoles, and cell surface staining was undetectable (Fig. 2A). According to a proposed structural model (8), the NH₂-terminal domain abuts membrane-spanning domains as well as the "A domain," one of the major cytoplasmic domains proposed to have critical catalytic function; it therefore seems plausible that removal of 60 residues leads to exposed hydrophobic regions and a nonfunctional molecule that cannot support viability. Because cell surface delivery was not detected and 60N Pma1 cannot sustain cell viability (not shown), this mutant was not studied further.

The residues between NH₂-terminal 40 and 60 are predicted to form an -helix followed by a loop (8). Residues 41–52 (DDIDALIEELQS), when modeled as an -helical wheel, forms an amphipathic helix, with hydrophobic residues lying on one face of the helix and positively charged residues on the opposing face. This sequence is conserved in all fungal P-type ATPases, and has been proposed to play a role in cell surface delivery (33). Therefore, mutagenesis was performed to change four aspartate (Asp^39–42) and two glutamate residues (Glu^48–49) to alanine to test whether the charged face of the helix might mediate protein-protein interaction during protein sorting and intracellular transport; this mutant is called Pma1-4D2E/A. Fig. 2B shows that newly synthesized Pma1-4D2E/A is predominantly delivered to the cell surface, but localization at both the plasma membrane as well as in intracellular puncta was seen by indirect immunofluorescence (Fig. 2A). Pma1-4D2E/A has decreased stability at the plasma membrane as it undergoes endocytosis for vacuolar degradation (see Fig. 4); however, Pma1-4D2E accumulation in vacuoles was only obvious in pep4 cells, likely because degradation is efficient in PEP4⁺ cells (Fig. 2A). Pma1-4D2E/A is competent to sustain cell viability as the sole copy of Pma1, as revealed by plasmid shuffle experiment (see Fig. 6).

Both Pma130C and Pma140C Truncation Mutants Are Partially Degraded by ER-associated Degradation—Because of its accumulation at the ER, we tested whether HA-Pma140 is a substrate for ERAD by pulse-chase analysis. Cells were pulse-labeled with [³⁵S]cysteine and methionine, and chased for various times. By contrast with wild-type Pma1, which has remarkable stability (Ref. 12 and Fig. 4, A and B), 40C, detected by immunoprecipitation with anti-HA, is rapidly degraded (Fig. 3A). No stabilization was observed in pep4 or end4 cells, impaired in internalization from the plasma membrane (not shown), but newly synthesized HA-tagged 40C is stabilized in mutants of the ER-associated ubiquitin-conjugating enzymes Ubc6 and Ubc7, and the ER-associated ubiquitin ligases Doa10 and Hrd1 (21) (Fig. 3A). These results suggest that degradation of HA-tagged 40C Pma1 requires ERAD machinery. Immunoprecipitation with anti-Pma1 antibody, which recognizes both tagged and untagged 40C, also shows loss of newly synthesized Pma1 (Fig. 3A, bottom panel).

View larger version (69K): [in this window] [in a new window]   FIGURE 1. Intracellular and plasma membrane localization of COOH-terminal truncation mutants. A, indirect immunofluorescence localization. Cells were grown to mid-log phase at 25 °C and then fixed with formaldehyde, spheroplasted, and permeabilized for staining. Staining with a single primary antibody was visualized with a CY3-conjugated secondary antibody; dichlorotriazinylaminofluorescein- and Texas Red-conjugated secondary antibodies were used for staining with two primary antibodies. HA-tagged 30C Pma1 (SSX9-1b) and 40C Pma1 (SSX5-1A) are localized intracellularly and at the plasma membrane. Cells were stained with both polyclonal anti-Kar2 and monoclonal anti-HA, or anti-Pma1 polyclonal antibody alone. Staining with anti-HA and anti-Kar2 is coincident in 40 Pma1 cells (right middle panels). B, cell fractionation by Renografin density gradients. Yeast strains were grown at 25 °C. Cells were lysed and fractionated, as described under "Materials and Methods." Membranes from each fraction were pelleted and analyzed by SDS-PAGE and Western blot. Anti-Gas1 (triangle), anti-GDPase (open circle), and anti-Sec61 (diamond) were used to detect plasma membrane, Golgi, and ER membrane fractions. Steady-state distribution of HA-tagged wild-type and mutant Pma1 detected by anti-HA (square), 40C-Pma1 detected by anti-Pma1 monoclonal antibody (closed circle), and membrane markers in each fraction is plotted as a percentage of total across the gradient. HA-tagged wild-type Pma1 (SSY3) are predominantly in peak fractions 9 and 10, whereas Pma1 mutants, 30C and 40C, display varying degrees of intracellular accumulation.

View larger version (84K): [in this window] [in a new window]   FIGURE 2. Intracellular and plasma membrane localization of NH2-terminal Pma1 mutants. A, indirect immunofluorescence localization. Strains with HA-tagged wild-type Pma1 (KKY37), 40N (KKY39), and 60N (KKY40) under the control of GAL1 were shifted from raffinose-containing synthetic complete medium to galactose-containing medium to induce Pma1 synthesis. Wild-type (L3852) and pep4 (ACY67) cells with pMET-Pma1-4D2E/A were washed with water and shifted to methionine-free medium to derepress mutant Pma1. After a 2-h induction cells were fixed and processed for staining with anti-HA antibody. B, cell fractionation by Renografin density gradients. Distribution of newly synthesized Pma1 (detected by anti-HA) and membrane markers is plotted as a percentage of total. Triangle, square, and cross indicate Gas1, HA-Pma1, and alkaline phosphatase, respectively. Cells with HA-tagged wild-type Pma1 (pKK98) or HA-tagged mutant Pma1-4D2E/A (pKK103) under the control of MET3 were grown to mid-log at 30 °C in minimal medium with 600 µM methionine, and induced by resuspending in methionine-free medium for 2 h. Cells were then lysed for fractionation. The majority of newly synthesized Pma1 is delivered to the plasma membrane.

View larger version (80K): [in this window] [in a new window]   FIGURE 3. Decreased stability of COOH-terminal truncation mutants as revealed by pulse-chase analysis. Cells were grown in minimal medium overnight at 25 °C before pulse labeling with [35S]methionine and cysteine for 10 min at 25 °C and chasing for various times. Immunoprecipitation from lysate was normalized to acid-precipitable counts/min, and analyzed by SDS-PAGE and fluorography. A, a fraction of newly synthesized 40 is degraded by ERAD. Anti-HA immunoprecipitates were from pulse-labeled 40 pma1 (SSX5-1a), 40 pma1 doa10 hrd1 (SSY24), and 40 pma1 ubc6 ubc7 (SSY25) cells. Bottom panel, immunoprecipitation from 40 pma1 cells with anti-Pma1 antibody. B, a fraction of newly synthesized 30 Pma1 undergoes ERAD. Anti-HA immunoprecipitates were from 30 pma1 (SSX9-1b), 30 pep4 (SSY39), 30 end4 (SSY38), and 30 ubc6 ubc7 (SSY37) cells.

Conformational Condition of Wild-type and Mutant Pma1 and Its Effect on Cell Growth—The first step in ERAD is thought to involve conformational recognition of a misfolded protein (2). Because both 40C and 30C are at least partially ERAD substrates, the conformation of these mutants was assayed by limited tryptic digestion. Membrane fractions were incubated with trypsin for various times and tryptic fragments were visualized by Western blot. Previous work from the laboratory of Carolyn Slayman (35) has shown that tryptic cleavage occurs initially at the NH₂ terminus so it is possible to detect tryptic fragments with either anti-Pma1 or anti-HA antibody recognizing the COOH-terminal tag. Fig. 5A (top panel) shows 30C Pma1 has increased susceptibility to tryptic cleavage and a different digestion pattern from that of wild-type Pma1; blotting with anti-HA revealed a similar tryptic cleavage pattern as that seen with anti-Pma1 (not shown). These results suggest a conformational difference between 30C and wild-type Pma1. Although the preponderance of 30C Pma1 is properly targeted to the plasma membrane (Fig. 1B), it appears that its conformational difference is detected by multiple quality control mechanisms as 30C undergoes ERAD (Fig. 3B) as well as vacuolar delivery.

View larger version (38K): [in this window] [in a new window]   FIGURE 4. Trafficking pathway of NH2-terminal mutants: stabilization of Pma1-4D2E/A by preventing endocytosis and vacuolar proteolysis. A, pulse-chase analysis. Cells with GAL-HA-PMA1 (KKY37) and GAL-HA-pma1-40N (KKY39) were grown in synthetic complete medium without cysteine and methionine with 2% raffinose overnight before induction of Pma1 by resuspending cells in 2% galactose for 1 h. Cells were then pulse-labeled at 25 °C with Expre35S35S for 10 min and chased for various times. Immunoprecipitation from lysate was normalized to acid-precipitable counts/min, and analyzed by SDS-PAGE and fluorography. B and C, degradation of Pma1-4D2E/A and its stabilization in end3-1 (RH266-1D) and pep4 (ACY67) but not doa10 hrd1 (MHY1703) cells. Cells bearing pMET3-HA-pma1-4D2E/A (pKK103) were grown to mid-log in the presence of methionine (off), induced to express mutant Pma1 by washing with water and resuspending in methionine-free medium for 2 h (on), and chased by addition of 2 mM methionine for up to 4 h.

View larger version (55K): [in this window] [in a new window]   FIGURE 5. Conformational analysis of wild-type and mutant Pma1 by limited trypsinolysis. Strains were grown at 25 °C in synthetic complete medium. Lysate was prepared and total membranes were pelleted by centrifugation at 100,000 x g for 1 h. Membranes were resuspended in buffer and incubated with trypsin (1:25 trypsin:protein) at 30 °C for various times. Digestion was terminated by acid precipitation, and products were analyzed by Western blot with anti-HA and/or polyclonal anti-Pma1 antibody. A, comparison of COOH-terminal Pma1 mutants (SSX5-1a, SSX9-1b) with wild-type Pma1 (SSY3). B, comparison of NH2-terminal mutants with wild-type Pma1. 40N (KKY44) cells were grown in synthetic complete medium with galactose because the mutant is under the control of the GAL1 promoter and the sole copy of Pma1 in the cell; the corresponding wild-type strain is KKX24-1B. The Pma1-4D2E/A mutant (KKY94) is under the control of the PMA1 promoter and it is the sole copy of Pma1 in the cell; the corresponding wild-type control is KKY93. Western blots are with anti-Pma1 antibody.

Consistent with their severe conformational defects, cells with 30C Pma1 and Pma1-4D2E/A display temperature-sensitive growth (Fig. 6). Pma1-4D2E/A also has growth defects at 25 and 30 °C (Fig. 6).

View larger version (127K): [in this window] [in a new window]   FIGURE 6. Growth of wild-type and Pma1 mutants. Strains were grown on synthetic complete medium at 25, 30, and 37 °C. Strains are wild-type (F1105), WT-HA (SSY3), HA-WT (KKY93), 30C (SSX9-1b), 40C (SSX5-1a), and Pma1-4D2E/A (KKY94).

To confirm that the COOH-terminal domain is not essential for oligomer formation, co-immunoprecipitation was assayed using myc-tagged and HA-tagged Pma1 constructs. Fig. 7B, lane 2, shows Western blot detection of myc-tagged wild-type Pma1 in a non-denaturing immunoprecipitate of HA-tagged wild-type Pma1, in agreement with previous work (36). Fig. 7B, lane 3, shows 40C Pma1 (myc-tagged) co-immunoprecipitates with wild-type Pma1 (HA-tagged). 40C Pma1 has a faster electrophoretic mobility compared with full-length Pma1 (compare lane 2 with other lanes). 40C Pma1 (myc-tagged) associates with 40C Pma1 (HA-tagged) (Fig. 7B, lane 4). Control experiments show that anti-HA antibody cannot immunoprecipitate myc-tagged Pma1 or recognize myc-tagged Pma1 by Western blot (Fig. 7B, lane 1); nor can anti-myc recognize HA-tagged Pma1 by Western blot (lane 3, upper band). These data suggest that the COOH-terminal domain is not essential for Pma1 oligomerization.

View larger version (67K): [in this window] [in a new window]   FIGURE 7. Pma1 mutants are not impaired in oligomerization. A, fast protein liquid chromatography analysis. Cell lysate was prepared from strains grown to mid-log in synthetic complete medium at 25 °C. Lysate was solubilized with 1% digitonin for 45 min at 25 °C. After removal of nonsolubilized material by centrifugation at 100,000 x g for 1 h, samples were fractionated by fast protein liquid chromatography using a column with a sizing range of 104 to 1.5 x 106. Protein was acid-precipitated and analyzed by Western blot with anti-HA and/or anti-Pma1 antibody. Analyzed strains (from top to bottom) are: 30C (SSX9-1b), 40C (SSX5-1A), HA-WT (KKY93), and 4D2E/A (KKY94). Migration of markers, blue dextran (void), aldolase (150 kDa), thyroglobulin (669 kDa), ferritin (440 kDa), and -amylase (200 kDa) are indicated by arrows. B, co-immunoprecipitation (IP). Cells were grown at 25 °C in synthetic complete medium with 2% raffinose. Cells were harvested and resuspended in synthetic complete medium with 2% galactose for 4 h. Lysate was prepared from wild-type cells bearing (1) GAL-myc-40pma1 only (pSS16), (2) both GAL-myc-PMA1 (pWQ3) and GAL-HA-PMA1 (pND542), (3) both pSS16 and pND542, and (4) both pSS16 and GAL-HA-40pma1 (pSS17). Immunoprecipitations (normalized to lysate protein) were with monoclonal anti-HA antibody and protein G-agarose beads in the presence of 1% Nonidet P-40, and analyzed by Western blot with polyclonal anti-myc antibody; lysates were examined by Western blot with both anti-myc and anti-HA antibodies.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Although cell surface delivery occurs, removal of the last 30 residues affects Pma1 conformation (Fig. 5), and trafficking. Some 30C Pma1 molecules are delivered for ERAD (Fig. 3), and 30C displays additional impaired trafficking in the secretory pathway, as revealed by localization in puncta and in the vacuole (Fig. 1A). The puncta may represent a slow step in 30C trafficking through endosomes or Golgi. It is not clear why 30C is delivered to multiple different transport pathways; however, we have previously observed that misfolded Pma1 mutants can escape ERAD but are then detected by quality control steps in other parts of the secretory pathway (1). It seems reasonable to propose that different quality control sites independently collect information about a discrete domain rather than assessing a global conformational state. It is also possible that conformational changes occur during intracellular transport so that a mutant protein can assume certain requisite conformations but not others in distinct organelles. In this way, a mutant protein can escape entirely or partially one quality control checkpoint, but undergo detection at another.

Removal of 40 residues from the COOH terminus resulted in two distinct populations of molecules: one population bearing the HA tag is retained for ERAD, whereas a second population is correctly targeted to the plasma membrane. Because no significant ER peak of 40C was seen by Western blot with anti-Pma1 antibody after cell fractionation (Fig. 1B), it appears that the ER-retained form might represent a very small fraction of total 40C Pma1. By contrast, the plasma membrane population is not well recognized by Western blot with anti-HA antibody (Fig. 1B); the simplest hypothesis to explain these results is that Pma1 that is delivered to the plasma membrane has lost the tag. The plasma membrane form of 40C appears stable. Interestingly, the HA-tagged 40 Pma1, which undergoes ERAD, has increased sensitivity to tryptic digestion, although it is not clear how the tag changes the conformational state of the protein.

Based on physical evidence from other family members, a structural model proposes that oligomerization of Pma1 is mediated by COOH-terminal interactions (5). Oligomerization of newly synthesized Pma1 is initiated in the ER (11). Even so, several observations support the idea that oligomerization is not required for Pma1 activity or its trafficking to the plasma membrane: Pma1 has catalytic activity when reconstituted as a monomer in liposomes (38, 39); Pma1 is a monomer at 30 °C in lcb1-100 cells, defective in serine palmitoyltransferase activity and sphingolipid synthesis (40), and yet these cells are viable at this temperature, suggesting monomeric Pma1 at the plasma membrane is functional (11, 36). Using our COOH-terminal truncation mutants, we tested the structural model by gel filtration chromatography as well as co-immunoprecipitation to examine oligomerization. Our results indicate that COOH-terminal truncation does not prevent Pma1 oligomerization. We suggest that oligomerization is not exclusively dependent on the COOH-terminal domain but may involve multiple contacts between monomers. Consistent with this idea, 30C, 40C, and Pma1-4D2E/A mutants are conformationally abnormal and yet are able to multimerize (Fig. 7). In addition, we showed previously that the ERAD substrate Pma1-D378N has a dominant negative effect on cell growth because it captures wild-type Pma1 for ERAD by forming hetero-oligomers with it; in the absence of Pma1 oligomerization in lcb1-100 cells, Pma1-D378N is suppressed (36). In this regard, it is revealing that a genetic screen to identify Pma1 mutants that suppress Pma1-D378N in trans did not yield any oligomerization-defective mutants,³ supporting the idea that numerous interactions participate in oligomer formation.

By contrast with COOH-terminal truncation mutants, the NH₂-terminal Pma1-4D2E/A mutant is not detected by ERAD. Indeed, Pma1-4D2E/A has no detectable targeting defect (Fig. 2) but has impaired stability at the plasma membrane (Fig. 4). Our results also suggest increased cell surface turnover of 40N. Our results are consistent with previous studies that concluded that the NH₂-terminal domain has little direct effect on overall enzyme function (33, 41). The diminished plasma membrane stability of the Pma1-4D2E/A mutant is like that of Pma1–10, which has point mutations in the cytoplasmic loop between transmembrane domains 2 and 3, comprising the A domain thought to play an important role in catalysis (13, 27). Because the cytoplasmic NH₂-terminal domain is proposed to abut the A domain (5), it seems possible that perturbations in either domain could affect protein-protein interactions that promote plasma membrane stability. Pma1–10 has a slow ER export phenotype that was detected by indirect immunofluorescence after inducing expression of the tagged protein (27). Using the same method in this work (Fig. 2A), no effect was detected on ER export of NH₂-terminal mutants. Thus, it seems unlikely that the NH₂-terminal domain carries an ER export motif.

Further work is necessary to understand whether the COOH-terminal domain of Pma1 carries an ER export signal or how some 30C and 40C molecules escape ERAD. Cells with 30C Pma1 as the sole Pma1 cannot grow at 37 °C; Pma1-4D2E also confers temperature sensitivity (Fig. 6). One possible explanation for temperature-sensitive growth is that there is increased ERAD and/or turnover from the cell surface to the vacuole at 37 °C, resulting in limited plasma membrane protein and activity at the cell surface. The temperature-sensitive phenotype provides the opportunity to use a genetic screen to gain further insight into these mutants.

	FOOTNOTES

 
^* This work was supported by National Institutes of Health Grant GM 58212. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

¹ To whom correspondence should be addressed: 830 N. University, Ann Arbor, MI 48109. Tel.: 734-647-7963; Fax: 734-647-0884; E-mail: amychang{at}umich.edu.

² The abbreviations used are: ER, endoplasmic reticulum; ERAD, endoplasmic reticulum-associated degradation; HA, hemagglutinin.

³ R. Tanaka, unpublished data.

	ACKNOWLEDGMENTS

 
We thank Randy Schekman, Howard Riezman, Greg Payne, Mark Hochstrasser, Charlie Boone for strains, antibodies, and plasmids, and Charles Yocum and Hana Popelkova for sharing their FPLC expertise and machine.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 

1. Arvan, P., Zhao, X., Ramos-Castaneda, J., and Chang, A. (2002) Traffic 3, 771–780[CrossRef][Medline] [Order article via Infotrieve]
2. Sayeed, A., and Ng, D. T. W. (2005) Crit. Rev. Biochem. Mol. Biol. 40, 75–91[CrossRef][Medline] [Order article via Infotrieve]
3. Sharma, M., Pampinella, F., Nemes, C., Benharouga, M., So, J., Du, K., Bache, K. G., Papsin, B., Zerangue, N., Stenmark, H., and Lukacs, G. L. (2004) J. Cell Biol. 164, 923–933[Abstract/Free Full Text]
4. Morsomme, P., Slayman, C. W., and Goffeau, A. (2000) Biochim. Biophys. Acta 1469, 133–157[Medline] [Order article via Infotrieve]
5. Kuhlbrandt, W. (2004) Nat. Rev. Mol. Cell. Biol. 5, 282–295[CrossRef][Medline] [Order article via Infotrieve]
6. Toyoshima, C., Nakasako, M., Nomura, H., and Ogawa, H. (2000) Nat. Rev. Mol. Cell. Biol. 405, 647–655
7. Auer, M., Scarborough, G. A., and Kuhlbrandt, W. (1998) Nature 392, 840–843[CrossRef][Medline] [Order article via Infotrieve]
8. Kuhlbrandt, W., Zeelen, J., and Dietrich, J. (2002) Science 297, 1692–1696[Abstract/Free Full Text]
9. Wurtele, M., Jelich-Ottmann, C., Wittinghofer, A., and Oecking, C. (2003) EMBO J. 22, 987–994[CrossRef][Medline] [Order article via Infotrieve]
10. Portillo, F. (2000) Biochim. Biophys. Acta 1469, 31–42[Medline] [Order article via Infotrieve]
11. Lee, M. C. S., Hamamoto, S., and Schekman, R. (2002) J. Biol. Chem. 277, 22395–22401[Abstract/Free Full Text]
12. Benito, B., Moreno, E., and Lagunas, R. (1991) Biochim. Biophys. Acta 1063, 265–268[Medline] [Order article via Infotrieve]
13. Gong, X., and Chang, A. (2001) Proc. Natl. Acad. Sci. U. S. A. 98, 9104–9109[Abstract/Free Full Text]
14. Barlowe, C. (2003) Trends Cell Biol. 13, 295–300[CrossRef][Medline] [Order article via Infotrieve]
15. Epping, E., and Moye-Rowley, W. S. (2002) J. Biol. Chem. 277, 34680–34869
16. Michelsen, K., Yuan, H., and Schwappach, B. (2005) EMBO Rep. 6, 717–722[CrossRef][Medline] [Order article via Infotrieve]
17. Sherman, F., Hicks, J. B., and Fink, G. R. (1986) Methods in Yeast Genetics: A Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY
18. Gietz, D., St. Jean, A., Woods, R. A., and Schiestl, R. H. (1992) Nucleic Acids Res. 20, 1425[Free Full Text]
19. Longtine, M. S., McKenzie, A., Demarini, D. J., Shah, N. G., Wach, A., Brachat, A., Philippsen, P., and Pringle, J. R. (1998) Yeast 14, 953–961[CrossRef][Medline] [Order article via Infotrieve]
20. Nothwehr, S. F., Bryant, N. J., and Stevens, T. H. (1996) Mol. Cell. Biol. 16, 2700–2707[Abstract]
21. Swanson, R., Locher, M., and Hochstrasser, M. (2001) Genes Dev. 15, 2660–2674[Abstract/Free Full Text]
22. Cross, F. (1997) Yeast 13, 647–653[CrossRef][Medline] [Order article via Infotrieve]
23. Goldstein, A. L., and McCusker, J. H. (1999) Yeast 15, 1541–1553[CrossRef][Medline] [Order article via Infotrieve]
24. Boeke, J. D., Trueheart, J., Natsoulis, G., and Fink, G. R. (1987) Methods Enzymol. 154, 164–175[Medline] [Order article via Infotrieve]
25. Raths, S., Rohrer, J., Crausaz, F., and Riezman, H. (1993) J. Cell Biol. 120, 55–65[Abstract/Free Full Text]
26. Sikorski, R. S., and Hieter, P. (1989) Genetics 122, 19–27[Abstract/Free Full Text]
27. Liu, Y., and Chang, A. (2006) J. Cell Sci. 119, 360–369[Abstract/Free Full Text]
28. Roth, A. F., Sullivan, D. M., and Davis, N. G. (1998) J. Cell Biol. 142, 949–961[Abstract/Free Full Text]
29. Rose, M. D., Winston, F., and Hieter, P. (1990) Methods in Yeast Genetics: A Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY
30. Chang, A., and Slayman, C. W. (1991) J. Cell Biol. 115, 289–295[Abstract/Free Full Text]
31. Chang, A. (2002) Methods Enzymol. 351, 339–350[Medline] [Order article via Infotrieve]
32. Mandala, S. M., and Slayman, C. W. (1988) J. Biol. Chem. 263, 15122–15128[Abstract/Free Full Text]
33. Portillo, F., de Larrinoa, I. F., and Serrano, R. (1989) FEBS Lett. 247, 381–385[CrossRef][Medline] [Order article via Infotrieve]
34. Jones, E. (1991) J. Biol. Chem. 266, 7963–7966[Free Full Text]
35. Lecchi, S., Allen, K. E., Pardo, J. P., Mason, A. B., and Slayman, C. W. (2005) Biochem. 44, 16624–16632[CrossRef][Medline] [Order article via Infotrieve]
36. Wang, Q., and Chang, A. (2002) Proc. Natl. Acad. Sci. U. S. A. 99, 12853–12858[Abstract/Free Full Text]
37. Mason, A. B., Allen, K. E., and Slayman, C. W. (2006) J. Biol. Chem. 280, 23887–23898
38. Goormaghtigh, E., Chadwick, C., and Scarborough, G. A. (1986) J. Biol. Chem. 261, 7466–7471[Abstract/Free Full Text]
39. Friant, S., Meier, K. D., and Riezman, H. (2003) EMBO J. 22, 3783–3791[CrossRef][Medline] [Order article via Infotrieve]
40. Zanolari, B., Friant, S., Funato, K., Sutterlin, C., Stevenson, B. J., and Riezman, H. (2000) EMBO J. 19, 2824–2833[CrossRef][Medline] [Order article via Infotrieve]
41. Hasper, A., Soteropoulos, P., and Perlin, D. S. (1999) Biochim. Biophys. Acta 1420, 214–222[Medline] [Order article via Infotrieve]

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This Article Abstract Full Text (PDF) All Versions