![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
Vol. 16, Issue 6, 2681-2693, June 2005
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
School of Biological Sciences, University of Nebraska–Lincoln, Lincoln, NE 68588-0118
Submitted June 12, 2004;
Revised March 2, 2005;
Accepted March 15, 2005
Monitoring Editor: Peter Devreotes
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
; Loomis, 1988; Brar and Siu, 1993), gp80 (Muller and Gerisch, 1978; Noegel et al., 1986), and gp150 (Gao et al., 1992; Wang et al., 2000), are mediating these interactions (reviewed in Kessin, 2001 and Siu et al., 2004). An understanding of the plasma membrane molecules involved in adhesive events during phagocytosis and motility of vegetative amoebae is emerging with the recent identification of cell-substrate adhesion molecule, sadA (Fey et al., 2002) and the phg1 family of transmembrane 9 proteins (Cornillon et al., 2000; Benghezal et al., 2003).
In this study, we present evidence that a plasma membrane glycoprotein gp130 also played a role in cell-substrate adhesion during vegetative growth. Postulated to be a phagocytosis receptor (Chia, 1996), gp130 is possibly the same molecule as gp126, a surface-exposed glycoprotein suggested to have a dual role as both a phagocytosis receptor and a mediator of cell-cell cohesion (Chadwick and Garrod, 1983; Chadwick et al., 1984; Chadwick, 1986). Characteristics of gp130 consistent with a role in phagocytosis are its presence in phagosomes, an association with detergent (Triton X-100)-insoluble cytoskeletons of bacterially grown cells but a depletion from membranes because of its internalization during phagocytosis (Chia, 1996; Rezabek et al., 1997), and the presence of an altered form in a D. discoideum phagocytosis mutant (Vogel et al., 1980). The relationship of gp126 to gp130 remains unresolved as the antibodies used for the studies on gp126 appeared to recognize carbohydrate rather than peptide epitopes. Antibodies raised against partially deglycosylated gp130 inhibited phagocytosis, but they too were not protein-specific as they recognized glycosylation modifications shared by D. discoideum glycoproteins (Chia and Luna, 1989). This finding supported the idea that carbohydrates could be partially responsible for cell-cell and cell-substrate interactions in D. discoideum but uncertainty remained regarding the function of gp130.
We report here the genomic sequence of the gene for gp130, expression patterns of the protein during growth and development, and biochemical and functional analyses of two transformed cell lines that specifically lack the protein. Gp130 was related to the family of gp138 proteins in D. discoideum that were postulated to participate in sexual cell fusion events leading to the formation of macrocysts (Aiba et al., 1993; Fang et al., 1993). As expected, gp130 was found at the cell-surface and expressed at high levels during vegetative growth of cells in axenic media. Protein levels were relatively low during growth on bacteria, a finding that undercut an essential role for gp130 in phagocytosis. Cells with the gene for gp130 disrupted by the UMP synthase gene cassette were healthy, growing faster and to higher densities than parent cells. They also were competent in macropinocytosis and phagocytosis, and made normal but slightly smaller fruiting bodies. Compared with parent cells and consistent with its plasma membrane location, gp130-null cells had altered surface properties. They displayed weaker cell-substrate adhesion but paradoxically were more adhesive with each other. The complex phenotype of the gp130-null cells pointed to roles for gp130 in plasma membrane trafficking and cell-substrate adhesion.
Preliminary findings and the sequence of gp130 was presented in at the 40th Annual meeting of the American Society for Cell Biology (LaRosa et al., 2000). The sequence is GenBank Accession No. AY038935
[GenBank]
.1 and also can be accessed through the Dictyostelium genome sequencing web site (http://www.dictybase.org) using DictyBase ID DDB0214937.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
; in the presence of 20 µg/ml uracil), transformants 1C7 and 2F5 (this study), and gp138-null (Hata et al., 2001; in the presence of 2 µg/ml blasticidin; Invitrogen, Carlsbad, CA) were grown at 22°C either axenically in HL5 medium (Watts and Ashworth, 1970) or in suspension with Klebsiella aerogenes as described (Chia, 1996) and shaken at 150 rpm. Cells were counted using a hemacytometer. Bacterial suspensions (inoculated with 1 x 105 D. discoideum cells/ml) were monitored using an OD600 of 1.0 corresponding to 1 x 109 bacteria/ml. Growth of cells on bacteria was unaffected by the inclusion or exclusion of uracil. For growth on lawns, K. aerogenes was grown overnight on SM plates at 37°C before inoculation with 1 x 105 D. discoideum cells (Sussman, 1987). Plates were put at 22°C and colony growth was measured daily. Three trials of growth were done, each with duplicates, and representative data are presented.
For starvation-induced development, cells were grown to midlog (5 x 106 cells/ml) before their collection by centrifugation. After two washes with Sorensen's buffer (SB; 14.6 mM KH2PO4, 2 mM Na2HPO4, pH 6.1), Resuspended cells (1 x 108 cells) were spread evenly on nonnutrient agar (1.5%) plates (100-mm diameter). Plates were stored in the dark and, over 24–36 h, images were recorded with a digital camera (Nikon Coolpix 995, Melville, NY) attached to a stereomicroscope (Nikon SMZ2800). Mounds, slugs, culminates, and fruiting bodies were counted at 12, 18, 24, and 36 h, respectively. For expression studies, cells scraped from plate surfaces were transferred to 1.5-ml microfuge tubes containing either 1 ml TRIzol reagent (Invitrogen, Carlsbad, CA) for RNA extraction or Laemmli sample buffer (Laemmli, 1970) containing 2 mM phenylmethylsulfonyl fluoride and a protease inhibitor mix (Roche Diagnostics, Indianapolis, IN) for protein analyses.
Nucleic Acid Analyses, Gene Disruption Construct and Transformation
A gene replacement vector was constructed by inserting the 
4-kb UMP synthase gene cassette containing the Dictyostelium pyr5-6 gene (DDPYR56G) into the unique ClaI site of the genomic sequence of gp130 cloned into pCR2.1 (Invitrogen). The cloning strategy used to acquire the gene for gp130 and details regarding its genomic sequence are described separately (Supplementary Data). The UMP synthase gene cassette had been released by digestion with ClaI from a gene replacement vector provided by Dr. M. Fechheimer (University of Georgia-Athens; Rivero et al., 1996) and was originally derived from plasmid pJB1 (Jacquet et al., 1988). The disrupted gp130 sequence was cut out of the pCR2.1 vector with sequential digestions using NotI and BamHI, and used to transform DH1 cells by electroporation (Knecht and Pang, 1995). Transformants were selected by their growth in FM media (Qbiogene, Carlsbad, CA) in the absence of uracil and subcloned by limiting dilution. Two transformed cell lines, 1C7 and 2F5, were analyzed in parallel in the studies presented.
Genomic DNA was isolated from cultures of DH1, 1C7, and 2F5 grown in HL5 to densities of 6–10 x 106 cells/ml. Cells were harvested and lysed to obtain nuclei that (Myre and O'Day, 2002) were lysed with Buffer G2 (Blood and Cell culture DNA Maxi Kit; Qiagen, Valencia, CA), and the DNA was further purified following to the protocol provided by the vendor. Using standard molecular techniques (Sambrook and Russell, 2001) DNA was digested with restriction enzymes for 24 h at 37°, precipitated, loaded, and separated on 0.7% agarose gels that were blotted to positively charged nylon membranes (Roche Diagnostics). DNA was quantified with the Hoescht assay (Sambrook and Russell, 2001), using calf thymus DNA (Sigma, St. Louis, MO) as a standard.
Probes for Southern blots were prepared by PCR using digoxigenin-11-UTP (alkali-labile DIG; Roche) at one part of the DIG-UTP synthesis mix to three parts of the standard dNTP stock for the gp130-specific probe. The DIG-UTP synthesis mix was used without dilution for the DDPYR56G probe. Primers for the gp130-specific probe (1457 base pairs) were 5'-CAACGGACCCATGTTTAGATAAT (P1; forward) and 5'-ATCTTTACCTTTAATAGTAATAATAG-3' (P2; reverse). Primers for the DDPYR56G probe (769 base pairs) were 5'-AGTAACAAGTGGTGCAAGTG-3' (P3; forward) and 5'-ACCAACACACAAAGAACC-3' (P4; reverse). The relative positions of primers P1–P4 are shown schematically in Figure 3A. Blots were processed according to the guidelines provided by Roche. Hybridization with the DDPYR56G-specific probe was at 42°C for 16 h. After exposure to film, blots were stripped according to Roche instructions and then hybridized with the gp130-specific probed at 37°C for 16 h. Exposures were 15 to 30 min.
|
). Using the gp130 cDNA template, [
-32P]dCTP was incorporated into DNA hybridization probes (1.3 kb) either by random priming using the Klenow fragment of DNA polymerase or gene specific primers and PCR. The blot was washed (each for 15 min at room temperature), first with 2x SSC, 0.1% SDS, followed by 0.5x SSC, 0.1% SDS, and then 0.2x SSC, 0.1% SDS and air-dried before exposure to film in the presence of intensifying screens.
Production of Bacterially Expressed Protein and Polyclonal Antiserum
After identifying a partial gp130 cDNA sequence, a translational fusion was constructed within the PQE-30 expression vector cassette (Qiagen) using BamHI (AACTTGGATCCACATTTATTAATGTTCCAATTCCA) and HindIII (ATATGAAGCTTAGATAAAATGATTAATGATAATAA) adapter primers for the 5' and 3' ends, respectively. The construct was electroporated into Escherichia coli strain M15 containing the pREP4 plasmid. Screening of transformants, grown at 37°C, expressing the fusion protein initially was performed by colony hybridization with a 32P labeled probes and later by immunoblotting using rabbit antibodies against a synthetic eight-residue peptide from gp130 (Chia, 1996). The fusion protein, predicted to have a molecular weight of 68.5 kDa, was comprised of an amino terminal sequence of MRGSHHHHHHGS followed by 609 amino acids (Thr151 to Ser759) of the conceptual preproprotein sequence derived from the cDNA. Restriction enzyme digests, PCR and cross-reactivity of the recombinant fusion protein to antipolyhistidine antibodies (His-probe H-15, Santa Cruz Biotechnology, Santa Cruz, CA) confirmed the production of the correct chimeric construct. Milligram quantities of the His-tagged protein were isolated and purified under denaturing conditions using nickel-nitrilotriacetic acid metal-affinity chromatography following vendor protocols (Qiagen). Antisera to the purified fusion protein were produced in rabbits at the Antibody Core Research Facility (Center for Biotechnology, University of Nebraska) following a standard immunization regime. An enriched IgG fraction that we call anti-gp130 was prepared from the antiserum by ammonium sulfate precipitation (Harlow and Lane, 1988) and used for detection of gp130 on protein blots.
Protein Analyses
Cells for lysates and plasma membrane preparations were harvested at densities between 6 and 9 x 106 cells/ml, unless otherwise noted in the figure legends. For lysates, cells were collected by centrifugation (200 x g) and washed twice in SB before lysis in 0.5% SDS at 80°C. Plasma membranes (PMs), were prepared by filter lysis and sucrose density centrifugation (Das and Henderson, 1983) from axenically grown Ax2 or gp138-null cells kindly provided by Dr. Urushihara (University of Tsukuba, Japan; Hata et al., 2001). Protein levels were measured using either the Bradford method (Bradford, 1976), after precipitating interfering SDS with 100 mM potassium phosphate, pH 7.2 (Zaman and Verwilghen 1979), or the Lowry method (Lowry et al., 1951) in the presence of 0.1% SDS and bovine serum albumin (Pierce Chemical, Rockford, IL) as a standard.
SDS-PAGE was performed essentially as described (Laemmli, 1970). Gels electrophoretically transferred to nitrocellulose (Towbin et al., 1979). Prestained markers (Benchmark, Invitrogen) were used to monitor electrophoresis. Blots were probed following standard protocols (Harlow and Lane, 1988) with antibodies followed by goat anti-rabbit or anti-mouse alkaline phosphatase conjugates (Southern Biotechnology Associates, Birmingham, AL). To detect glycoproteins, blots were probed with biotinylated concanavalin A (Con A; Biomeda, Foster City, CA) followed by streptavidin-alkaline phosphatase (Pierce Chemical; Chia, 1996). A previously described rabbit antibody was raised against a synthetic eight-residue peptide of gp130 (Chia, 1996). For this work, a different rabbit antibody (anti-gp130) was raised against bacterially expressed recombinant truncated version of gp130 (see above). Mouse monoclonal antibody AC11 reactive with gp138 was provided generously by Dr. Urushihara (Aiba et al., 1993).
For immunolocalization, log-phase axenic cells were fixed and permeabilized at –10°C with 1% formaldehyde in methanol using the agar-overlay method (Fukui et al., 1987). Slides were stained with or without anti-gp130 (75 µg/ml), followed by either FITC conjugated to goat anti-rabbit IgG (Sigma) adsorbed against glutaraldehyde-fixed Ax2 (Clarke et al., 1987) or Cy2 donkey anti-rabbit conjugates (Jackson ImmunoResearch Laboratories, West Grove, PA). Coverslips were mounted in buffered Gelvatol containing 50 mg/ml 1,4-diazobicyclo-(2,2,2)-octane (Aldrich Chemical, Milwaukee, WI). Images were collected with a Bio-Rad MRC1024ES confocal laser-scanning microscope (Richmond, CA) at the University of Nebraska Center for Biotechnology Microscopy Core Facility.
Assays of Function
Fluid-phase Endocytosis, Exocytosis, and Phagocytosis. Macropinocytosis and exocytosis were measured with the fluid-phase marker FITC-dextran (Mr 70,000; Sigma; Seastone et al., 2001). For endocytosis, cells in midlog phase were diluted with HL5 to 3 x 106 cells/ml and shaken with 2 mg/ml FITC dextran for 3 h. At various times cells were harvested, washed twice with SB, and then lysed with 0.5% Triton X-100 in SB. The fluorescence was measured with a Cary Eclipse spectrofluorimeter (Varian Instruments, Walnut Creek, CA) using 492 and 525 nm as the excitation and emission wavelengths, respectively. Duplicate samples of all cells were run in at least two trials of each assay. Live cells were imaged with an Olympus FluoView 500 (Olympus, Melville, NY) laser scanning microscope equipped with a 60x water immersion objective lens. Volumes of cell pellets were determined by centrifuging (2, 5, or 10 x 106) cells into calibrated micro-(capillary) pipettes. For exocytosis, cells in midlog phase were diluted with HL5 media to 3 x 106 cells/ml and loaded with 2 mg/ml FITC-dextran for 3 h. Cells then were washed twice with cold HL5 and resuspended in fresh HL5 without FITC-dextran. At times over a 3-h period, cells were harvested, washed twice with cold SB, and lysed with 0.5% Triton X-100 in SB, and their fluorescence was determined. To calculate the percent of fluorescence remaining in the cells, the fluorescence value at each time point was compared with the initial fluorescence, which was assigned a value of 100%. Phagocytosis was measured using Ds-Red fluorescent E. coli (Maselli et al., 2002).
Motility and Adhesion
Cells in midlog phase were diluted with HL5 to 3 x 106 cells/ml, and 300 µl was placed in each well of an eight-well chambered slide (LabTek, Nagle Nunc International, Naperville, IL). For starved cells, midlog phase cells were harvested, washed twice with cold SB, and resuspended to 3 x 106 cells/ml, and 300 µl was left in a well for 5 h before recording images (Yuen et al., 1995). An inverted microscope (Nikon Diaphot-TMD) equipped with Hoffman Modulation Contrast optics (40x objective; Modulation Optics, Greenvale, NY), camera (model DC-330 CCD; DAGE-MTI, Michigan City, IN) and CG-7 frame grabber (Scion, Frederick, MD) was used to collect video images at 1 frame/s in QuickTime 6.03. Three hundred frames were recorded and cell movement was measured by marking the position of the center of 25 cells every 60 frames. The total distance the cell traveled over a 5 min was determined and converted after determining the magnification using a micrometer scale observed with the same lens.
Cell-cell adhesion was assayed using a method previously described (Desbarats et al., 1994). Midlog phase cells were collected and diluted to 3 x 106 cells/ml in HL5 (vegetative conditions) or SB (starvation conditions) and shaken at 22°C. At times indicated, 500 µl were transferred to a 1.5-ml microfuge tube, vortexed for 20 s, and mixed for 20 min using a Labquake shaker (Barnstead/Thermolyne C400110). Adhesion was measured by counting the number of single cells, and the number of cells in aggregates with a hemacytometer at each time point. Doublets were counted as aggregated cells. For suspension-development trials, late log phase axenically grown cells were harvested, washed twice in SB, and then suspended to 1 x 107 cells/ml in 17 mM phosphate buffer (pH 6.4) ± 5 mM EDTA. Because of their documented behavior during suspension-development, Ax3 cells were used as control cells instead of DH1. Cells were rotated at 180 rpm on a platform shaker, and at indicated times, aliquots were diluted to 2 x 106 cells/ml in the same buffer. Unaggregated cells (singlets and doublets) were counted, and the percentage of aggregated cells was determined (Desbarats et al., 1994).
Cell substrate adhesion was measured following Barondes et al. (1987). Cells in midlog phase (5 x 106 cells/ml) was collected and diluted to 1 x 106 cells/ml in HL5. One hundred microliters of cells were added to wells of a 96-well polystyrene plate (Corning 25801) and allowed to settle and attach at room temperature for 30 min. The wells were then completely filled with HL5 without disturbing attached cells, sealed with an adhesive plastic sheet, inverted, and centrifuged for 5 min at 300 x g at room temperature in a rotor with swinging microplate holders. After draining unbound cells, 150 µl of 0.4% SDS in SB was added to adhering cells, and the plate was shaken on a plate shaker at 37°C for 15 min to ensure complete lysis. Then, 150 µl of undiluted Coomassie Protein Assay Reagent (Pierce Chemical) was added to each well, and the plate was again shaken for 15 min at 37°C. The absorbance at 595 nm was read on a VersaMax microplate reader (Molecular Devices, Sunnyvale, CA) running SoftMax ProVer 4.6. The percentage of cells remaining in the wells was calculated from a standard curve generated by determining the amount of protein of a known number of cells and lysed in parallel with the experimental samples. At least three separate sets of the assay were done, with each set having the two transformants and parent DH1 in duplicate on the same plate.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
). Seven unique peptide sequences acquired from proteolyzed, gel-purified gp130 were back-translated to generate oligonucleotides for PCR with different templates. Concurrently, a partial genomic clone containing two sequenced peptides was identified in the databases of the D. discoideum Sequencing Consortium. This genomic clone, together with products from PCR trials, primer walking and anchored PCR, yielded first a cDNA sequence and subsequently the full genomic sequence for gp130 (see Supplementary Data for details).
From the amino acid sequence deduced from the coding sequence (2307 bases; GenBank accession no. AY038935
[GenBank]
.1), the predicted precursor of gp130 has a mass of 85.3 kDa (Figure 1A). If the predicted loss of signal peptides both at the N-and C-termini is taken into account (see below), the mass of the mature protein would be 79.7 kDa, which is nearly 50 kDa less than the Mr of 130 kDa observed originally on SDS-polyacrylamide gradient gels (Chia, 1996). The disparity likely is due to N-glycosylation, detected previously (Chia, 1996) of some of the 15 asparagines found in the Asn-X-Ser/Thr sequence used to predict glycosylation (Bause, 1983). The added carbohydrate would increase the mass of the molecule and also could retard its migration in SDS-gels. Weak glycosylation sites (when X is D, W, or P) and a site (N232) in a sequenced peptide are not indicated. Eleven Cys residues were present in mature gp130 and the absence of reductant increased slightly its SDS-gel mobility compared with the reduced form (Schmaltz and Chia; unpublished data). The strong biotinylation of gp130 that demonstrated its cell-surface exposure (Chia, 1996) was attributable to the numerous lysines potentially available (39 in the mature protein), and perhaps amino sugars, to react with amine-directed modifying reagents.
|
; Aiba et al., 1997; Hata et al., 2001). When the protein sequence of gp130 was aligned with the gp138 proteins using ClustalW (unpublished data; Thompson et al., 1994; Accelrys, 2001), the position of the initial methionine of gp130 corresponded closely to the N-termini of the gp138 precursor proteins. Finally, using the assigned methionine, analyses of the preproprotein predicted a signal peptide (SPScan; Accelrys, 2001; and SignalP V1.1; Nielsen et al., 1997), consistent with the biosynthesis of gp130 via the secretory pathway and its location at the plasma membrane (Chia, 1996). Attempts to acquire an N-terminal sequence of gel-purified gp130 were unsuccessful as it appeared blocked, so the first residue of the mature protein could not be verified.
Prediction programs for protein secondary structure indicated two major hydrophobic stretches of gp130 that could be embedded in the membrane and identified several additional weakly hydrophobic short, internal segments. The possible membrane domains correspond to an N-terminal signal peptide of 21 residues and a 29-residue sequence at the C-terminus (Accelrys, 2001). No other transmembrane regions were predicted by the transmembrane hidden Markov model (TMHMM) routine (http://www.cbs.dtu.dk/services/TMHMM; Krogh et al., 2001). PSORT, a program to analyze protein localization, also predicts a 21-residue signal peptide (Nakai and Kanehisa, 1992). Similarly, the "big-II" routines (http://mendel.imp.univie.ac.at/gpi/gpi_server.html; Eisenhaber et al., 2000, 2001) indicated only two hydrophobic regions long enough to span a membrane. These correspond again to a signal peptide of 21 amino acids and the extremely hydrophobic 29-residue C-terminal sequence. The latter was predicted to direct the addition of a glycosylphosphatidylinositol (GPI) anchor. Previously, gp130 was postulated to be an intrinsic membrane protein, on the basis of its enrichment in purified plasma membrane fractions and a requirement of detergents for its solubility (Chia, 1996; Rezabek et al., 1997). A protein with a GPI-anchor would have similar properties. Other than two repeated peptides (FSLPS and SINVG), no other notable features or credible motifs in the protein were detected using various gene and protein analysis programs.
Southern blot hybridization patterns were consistent with a single copy of the gene for gp130 (unpublished data). This finding was confirmed by BLAST searches of the Dictyostelium genome sequence databases that revealed a single copy of the gene. There was however significant sequence similarity of gp130 to the D. discoideum gp138 family of cell-surface glycoproteins postulated to mediate sexual cell fusion (Fang et al., 1993; Aiba et al., 1997), although questions have been raised regarding this function (Hata et al., 2001). The gp138 genes map to chromosome 5 (Hata et al., 2001), whereas the gp130 gene is on chromosome 3. Pairwise comparisons, using BESTFIT and GAP (Accelrys, 2001) show that the four gp138 proteins were much more similar to each other (82–87% similarity and 77–85% identity) than to gp130. Comparison of gp130 to each of the gp138 proteins showed it to have 40–42% similarity and 34–35% identity with the group. Thus, gp130 is related to the family of gp138 proteins, and the five proteins together comprise a distinct group in D. discoideum with no known relatives in other eukaryotes.
Expression and Localization of gp130
The sequenced gene was verified to encode for gp130 when polyclonal antibodies, generated against a truncated version of gp130, reacted strongly with the recombinant protein and specifically with gp130 (Figure 1). The recombinant protein used as the antigen migrated at 65 kDa, close to its predicted molecular mass of 68.5 kDa (Figure 1B, lane 1). Gp130 appeared as two closely spaced species in lysates of vegetative cells (Figure 1B, lane 2). As expected, a strong signal for gp130 was present in purified plasma membranes prepared from log-phase Ax2 and gp138-null cells (Figure 1B, lanes 3 and 4). The gp130 doublet in plasma membranes was more evident with reduced protein loads (Figure 1E, lanes 2 and 3) and may be due to glycosylation differences. Anti-gp130 did not cross-react with gp138. The anti-gp130 signal in plasma membranes from gp138-null cells (Figure 1B, lane 4) migrated faster than anti-gp138 signals in Ax2 lysates and membranes (Figure 1C, lanes 2 and 3). The recombinant protein did not bind Con A, whereas the gp130 signals of lysates and plasma membranes corresponded to species that bound Con A (Figure 1D), indicating them to be glycoproteins. This observation supported the idea that glycosylation could account both for the doublet gp130 species and the slower mobility of the authentic protein. Deglycosylation of gp130 in plasma membranes using commercially available reagents was only partially effective, as judged by Con A blots of the treated plasma membranes (unpublished data). The reagents may have limited access to the glycosylated residues or the nature of the oligosaccharide modifications made them resistant to the treatments.
Gp130 was localized to the plasma membrane of axenically grown Ax2 cells stained with anti-gp130 as shown by the nearly exclusive staining of cell edges in confocal micrographs of fixed and permeabilized cells (Figure 2). This observation reinforced biochemical studies that showed gp130 to be readily surface-labeled with impermeant cell-labeling reagents and therefore found at the cell surface, and highly enriched in purified plasma membranes (Chia, 1996). Anti-gp130 reacted neither with native protein in plasma membranes nor Triton X-100–solubilized gp130 (as judged by ELISA analyses; unpublished data). It also did not bind to live cells, precluding its use in assays of live cells. The inability of anti-gp130 to react to native protein was expected since bacterially expressed protein was used as the antigen. Another contributing factor that could hinder its reactivity is that the polypeptide backbone of gp130 is masked by carbohydrate modifications, which were likely to be substantial based on the retarded mobility of gp130 on SDS-gels.
|
2.5 kb (Figure 2B). Levels of mRNA were high during vegetative growth and decreased during starvation-induced development. This pattern correlated well with the strong gp130 protein signal detected during (axenic) vegetative growth, whereas during development protein levels declined (Figure 2B). The lower levels of gp130 during development when cells are phagocytically inactive were consistent with the previous suggestion of a role for gp130 in phagocytosis (Chia, 1996; Rezabek et al., 1997). Unexpected, however, was the observation that during growth phase, levels of gp130 were lower in bacterially grown cells compared with axenically grown cells (Figure 2C, top). The loss of protein was reflected in the northern blots showing reduced mRNA levels in phagocytically active cells compared with cells grown in nutrient media (Figure 2C, bottom). An examination of gp130 levels by immunoblotting in D. discoideum strain NC4 also showed only modest levels of gp130 (unpublished data). These expression patterns of gp130 during vegetative growth indicated that the relative amount of gp130 was inversely correlated with the phagocytic activity of cells.
Disruption of the Gene for gp130
DH1 cells, auxotrophic for uracil (Caterina et al., 1994), were transformed with a construct where the UMP synthase gene cassette (Jacquet et al., 1988) was inserted into the genomic sequence of gp130 (Figure 3A). Transformed cells were selected by their ability to grow in FM without uracil. Southern blots of genomic DNA from transformants 1C7 and 2F5 were probed with either gp130-specific or UMP synthase-specific probes. The hybridization patterns were consistent with a specific disruption of the genomic gene for gp130 with the cassette (Figure 3B).
Immunoblot analyses also showed the absence of gp130 in the transformants (Figure 3C). A signal for gp130 was present in DH1 cell lysates and plasma membranes (PMs; lane 1 in panels i–iii) but missing in 1C7 and 2F5 (lanes 2 and 4, respectively, in panels i–iii). For comparison, the gp138-null cell line was analyzed in parallel. As shown in Figure 1C, these cells lack two gp138 signals immunoreactive with monoclonal AC11 (Aiba et al., 1993) that were present in PMs of DH1, 1C7, and 2F5 (panel iii). The Con A binding pattern of PMs from DH1, 2F5, and gp138 null cells showed the absence of a species corresponding to gp130 in 2F5 (panel iv, black circle). Also, compared with the parent DH1, there appeared to be an enhanced gp138 signal in 2F5 (lane 2, panel iii). This aligned with a Con A binding species present in PMs of 2F5, but absent in DH1 and gp138 null cells (panel iv, black square). Corroborating the lack of gp130 in PMs of the transformants, immunofluorescence microscopy showed 1C7 and 2F5 without the cell surface signal seen in DH1 cells (panel v). Together, the DNA and protein analyses support the conclusion that 1C7 and 2F5 were genuine transformants in which the gene for gp130 was disrupted.
Consequences of Disrupting the Gene for gp130 Growth and Development
The growth of the gp130-null cells appeared superior to DH1 both in suspension and on lawns of bacteria (Figure 4). In comparison to DH1 (generation time, GT, of 9 h), 1C7 and 2F5 grew faster (GT of 7–8 h) and reached higher densities in HL5 before reaching stationary phase (Figure 4A). Similarly, 1C7 and 2F5 exhibited faster growth rates (GT = 5 h) and achieved greater densities than DH1 (GT = 5.5 h) when grown in bacterial suspensions (Figure 4B). Neither the presence nor absence of uracil in DH1 suspension cultures had an effect on its growth (unpublished data). It seemed unlikely that gp130 was a receptor for specific nutrients present in HL5 because bacteria also were ingested at more rapid rates by the transformants. Both 1C7 and 2F5 also exceeded DH1 in terms of rate and overall size of colony growth on lawns of bacteria (Figure 4C). In addition, the clear zones of colonies of transformants expanded continuously over a 10-d period, whereas those of DH1 expanded slowly and ceased growing after a week (Figure 4, D and E). The absence of the gp130 protein appeared to enhance aspects of vegetative growth of the transformants.
|
|
|
Another altered membrane property of the transformants was revealed by adhesion assays. Tests of cell-substrate adhesion using plastic (polystyrene) surfaces showed that gp130-null cells were nearly 50% less adherent to the plastic than DH1 cells (Figure 7A). The decreased cell-substrate adhesion did not, however, lead to greater motility, as the average velocity of vegetative 1C7 and 2F5 cells was close to that of DH1 (Figure 7B). The total distance traveled by the cells in a 5-min period and the patterns of cell movement were comparable (unpublished data). After 5 h in buffer (starvation), the three cell lines uniformly exhibited greater motile activity and moved similar distances (unpublished data).
|
) was likely responsible for the cohesion of the starved cells because 5 mM EDTA decreased the percentage of aggregates of 2F5 and Ax3. Under these conditions, however, gp130-null cells formed more aggregates than Ax3, raising the suggestion that the absence of gp130 revealed an EDTA-insensitive cell-cell adhesion mechanism during early development. Cell lysis occurred when transformants were shaken in buffer with EDTA for more than 5 h, so these studies did not go beyond 6 h.
|
In summary, compared with parent DH1 cells, gp130-null cell lines 1C7 and 2F5 during vegetative growth had increased fluid-phase endocytosis, enhanced cell-cell cohesiveness, and decreased cell-substrate adhesion, whereas phagocytosis and motility appeared similar. The altered cell-surface properties appeared due to the absence of gp130 on the plasma membrane and indicated that gp130 influenced how D. discoideum cells interacted with each other and other surfaces. Because expression patterns of normal cells indicated a depletion of gp130 during starvation and development, predictably, under these conditions, gp130-null cells on substrates behaved largely like DH1 cells.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
). The proper reading frame and positions of two introns were established by seven internal microsequenced peptides. Additional corroboration of the sequence was aided by the identification of the sequence in contigs later available through the databases of the Dictyostelium genome sequencing project (Supplementary Data Figure 1). The presence and relative abundance of gp130 in vegetative (log-phase) axenically grown cells and purified plasma membranes, and its uniform plasma membrane location were consistent with a phagocytic function (Figures 1 and 2A). The amount of gp130 protein and its transcript were, however, unexpectedly lower in bacterially grown cells than in axenically grown cells (Figure 2, B and C). This expression pattern was inconsistent with a vital involvement of gp130 in phagocytosis. Further, because null cells were efficient phagocytes (Figures 4B and 6D), the protein was dispensable for phagocytosis, although different particles and solution conditions were not tested.
Aside from an expected N-terminal signal sequence, a C-terminal hydrophobic sequence that might direct the addition of a GPI-anchor, and N-glycosylation motifs, conserved protein regions or patterns in gp130 were not evident nor did various programs (Bateman et al., 2002; Hulo et al., 2004) that predict protein secondary structure provide immediate insights into the function of gp130. Additionally, gp130 had no significant resemblance to proteins in the databases except for the gp138 family suggested to be involved in sexual cell fusion of D. discoideum (Fang et al., 1993; Hata et al., 1999), although a subsequent study raises doubts about this role (Hata et al., 2001). The absence of relatives for gp130 and the gp138 group in other organisms suggests that these glycoproteins have unique roles in the biology of D. discoideum.
Function of gp130
Mutants lacking gp130 in D. discoideum were generated with the expectation that the cells would provide insights into the function of the protein. Southern blot analyses showed the gene for gp130 was disrupted, and immunoblots showed the corresponding absence of gp130 in transformants (Figure 3). Biochemical and physiological analyses indicated that cells lacking the protein had defects in neither growth nor developmental. Rather, gp130-null cells had faster growth rates and attained higher densities in nutrient HL5 media and bacterial suspensions, before reaching stationary phase, than the parent DH1 (Figure 4, A and B). The more robust growth of 1C7 and 2F5 in HL5 could be attributed to the higher activities of macropinocytosis (Figure 6A). Although particle uptake rates were similar for the cells when measured directly for short times (Figure 6D), growth of the gp130-null cells in bacterial suspensions was superior to DH1 (Figure 4B). The higher growth rates might be related to greater endocytotic activity that was more evident over time and consistent with the enhanced rates of fluid-phase uptake.
Studies of endocytosis in mammalian cell lines, yeast and D. discoideum commonly yield cases where macropinocytosis is disrupted rather than improved after gene mutations. For example, because the cytoskeleton plays a role in endocytosis (Qualmann et al., 2000; Qualmann and Kessels, 2002; Engqvist-Goldstein and Drubin, 2003), mutations in actin-binding proteins adversely affect macropinocytosis in D. discoideum (Maniak, 2001). A faster doubling time by D. discoideum cells lacking the substrate adhesion receptor sadA was due to the cytokinesis of multinucleate cells and not improved endocytosis (Fey et al., 2002). It seems unlikely that the improved growth characteristics of the gp130-null cells were a consequence of the inserted UMP synthase gene because a number of cell lines with genes disrupted with this cassette display rates of endocytosis the same as or less than that of the DH1 parent cell line (Rivero et al., 1996; Dragoi and O'Halloran, 1998; Liu et al., 2002). The robust growth characteristics of gp130-null cells thus were unforeseen. We offer two explanations for the observed behaviors of the transformants. One is that because gp130 was a component of the plasma membrane, its absence caused changes in the physical nature of the membrane bilayer that led to greater endocytotic activity. A second related scenario is that gp130 was involved directly in a mechanism governing membrane recycling. Its absence led to elevated rates of membrane trafficking, with the consequence of increasing overall endocytotic activity. Because cytoskeletal-membrane interactions are vital for macropinocytosis, the relationship of gp130 to other membrane or cytoskeletal proteins is a focus of future studies.
Colony growth rates as well as the continued expansion of clear zones on bacterial lawns by the gp130-null cells pointed to their faster movement on surfaces (Figure 4, C and D). This was consistent with their altered membrane flux and their decreased cell-substrate adhesion (Figure 7A) because cell migration is closely linked to interactions between cell and substratum (Bray, 1992; DeMali and Burridge, 2003). However, the rate of movement of the cells in HL5 on glass slides was similar to that of DH1 (Figure 7B), and their patterns of movement appeared the same when comparing their paths. The larger colony sizes and clear zones may reflect simply the enhanced growth rates of gp130-null cells seen in suspension culture.
Alternatively, the nature of the substrate may influence the strength of the cell-substrate interaction. This could explain how the weaker cell-substrate adhesion of the gp130-null cells allowed them to be fully capable of phagocytosis, a localized cell-substrate adhesion event (May and Machesky, 2001). Typically, cell-substrate adhesion and phagocytic competence are highly correlated (Bray, 1992; May and Machesky, 2001; DeMali and Burridge, 2003). One example is the sadA-null cells unable to attach to plastic that are also phagocytically incompetent (Fey et al., 2002). Likewise, phg1 cells, generated by insertional mutagenesis, have defective phagocytosis that coincides with faulty adhesion (Cornillon et al., 2000). Mutants of cortical cytoskeletal proteins such as talin (Niewöhner et al., 1997) and myosin VII (Tuxworth et al., 2001) similarly show a tight coupling of phagocytosis and adhesion. Because gp130-null cells had 50% of the cell-substrate adhesion of the parent DH1 cells, but were phagocytically functional, it appeared that gp130 participated in a cell-substrate adhesion mechanism different from the sadA and phg1 proteins, and also distinct from that used in the recognition of Klebsiella aerogenes and E. coli bacteria. We thus consider the idea that gp130 was a lectin-type receptor postulated by Vogel et al. (1980). Phagocytosis by the gp130-null cells could be performed by the "nonspecific" receptors that compensated for gp130 so its absence was not a liability. We also consider the possibility that gp130 was part of the cellular adhesion assemblage of surface molecules postulated to be controlled by the transmembrane 9 family of proteins Benghezal et al. (2003). Thus, gp130 could still have a role in phagocytosis as a specific receptor. Because the adhesion assay tested only cell interactions with plastic, the nature of a cell surface-adhesion system mediated by gp130 should be explored further by testing how cells ingest particles with different chemical properties under different conditions (Vogel et al., 1980; Cornillon et al., 2000).
Although the transformants exhibited weaker cell-substrate adhesion, it was paradoxical to find that compared with the cell-cell cohesion of DH1 cells, vegetative gp130-null cells were relatively more cohesive with each other in suspension (Figure 7C). Although cell-cohesion is typically not measured in vegetative cells, Marin et al. (1980) observed that nearly half the vegetative cells make aggregates (of three or more cells) within 30 min in a suspension of nutrient media.
The interaction of gp130-null cells was not wholly attributable to gp24, the molecule mediating calcium-dependent cell-cell cohesion during early development (Knecht et al., 1987; Loomis, 1988; Brar and Siu, 1993), because neither EDTA (Figure 8A) nor calcium (unpublished data) reduced the cohesion of transformants from log-phase cultures. When dis-aggregated and returned to HL5, transformants exhibited stronger cohesion than DH1 (Figure 8B). Suspension-starved transformants under higher shear forces also made more aggregates than control cells (Figure 8C). A component of this aggregation was apparently mediated by gp24 as the inclusion of EDTA decreased the total percentage of transformants and Ax3 cells in aggregates. Transformants, however, maintained a greater number of EDTA-resistant aggregates than Ax3 cells. These observations indicate that cells had the ability to self-cohere, in an EDTA-insensitive manner, which became more evident when gp130 was absent. Although the effectiveness of EDTA in HL5 is unclear, the adhesion mechanism displayed by vegetative transformants (Figure 8A) and the EDTA-resistant self-cohesion in shaken suspension were likely related. Studies of gp24-null cells revealed a class of EDTA-sensitive cell adhesion sites (Wong et al., 2002), although we do not know whether these are responsible for cohesive interactions of the vegetative transformants. Because gp130 contributed to substrate adhesion, a possibility is that its presence masked or suppressed cell-cell cohesion. Alternatively, the greater self-cohesion of transformants could be due to precocious expression of gp80, the EDTA-resistant cell-cell adhesion molecule. Desbarats et al. (1994) showed that cell-cell adhesion and contact are involved in the induction of gp80 expression.
The complex phenotype of gp130-null cells pointed to a role for gp130 in inhibiting cell-cell cohesion during vegetative growth while participating in cell-substrate adhesion when cells were in contact with dissimilar surfaces. Gp130-null cells also exhibited increased macropinocytosis, indicating that the absence of the glycoprotein enhanced membrane trafficking. We anticipate that analyses of cell lines that maintain or overexpress gp130 during development, or express fluorescently tagged gp130, will provide more details regarding its adhesion functions and impact on membrane flux.
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
|
Footnotes |
|---|
The online version of this article contains supplemental material at MBC Online (http://www.molbiolcell.org). 
Address correspondence to: Catherine P. Chia (cchia{at}unlserve.unl.edu).
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
Aiba, K., Fang, H., Yamaguchi, N., Tanaka, Y., and Urushihara, H. ((1997). ). Isoforms of gp138, a cell-fusion related protein in Dictyostelium discoideum. J. Biochem. 121, , 238–243.
Aiba, K., Yanagisawa, K., and Urushihara, H. ((1993). ). Distribution of gp138, a cell surface protein responsible for sexual cell fusion, among cellular slime moulds. J. Gen. Microbiol. 139, (Pt 2), 279–285.[Medline]
Barondes, S. H., Cooper, D. N., and Springer, W. R. ((1987). ). Discoidins I and II: endogenous lectins involved in cell-substratum adhesion and spore coat formation. Methods Cell Biol. 28, , 387–409.[Medline]
Bateman, A., Birney, E., Cerruti, L., Durbin, R., Etwiller, L., Eddy, S. R., Griffiths-Jones, S., Howe, K. L., Marshall, M., and Sonnhammer, E. L. ((2002). ). The Pfam protein families database. Nucleic Acids Res. 30, , 276–280.
Bause, E. ((1983). ). Structural requirements of N-glycosylation of proteins. Studies with proline peptides as conformational probes. Biochem. J. 209, , 331–336.[Medline]
Benghezal, M., Cornillon, S., Gebbie, L., Alibaud, L., Bruckert, F., Letourneur, F., and Cosson, P. ((2003). ). Synergistic control of cellular adhesion by transmembrane 9 proteins. Mol. Biol. Cell 14, , 2890–2899.
Bradford, M. ((1976). ). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principles of protein-dye binding. Anal. Biochem. 72, , 248–254.[CrossRef][Medline]
Brar, S. K., and Siu, C.-H. ((1993). ). Characterization of the cell adhesion molecule gp24 in Dictyostelium discoideum. J. Biol. Chem. 268, , 24902–24909.
Bray, D. ((1992). ). Cell movements, New York: Garland.
Caterina, M. J., Milne, J. L., and Devreotes, P. N. ((1994). ). Mutation of the third intracellular loop of the cAMP receptor, cAR1, of Dictyostelium yields mutants impaired in multiple signaling pathways. J. Biol. Chem. 269, , 1523–1532.
Chadwick, C. M. ((1986). ). Changes in the cell surface level of gp126 during development of Dictyostelium discoideum. Devel. Growth Differ. 28, , 203–211.[CrossRef]
Chadwick, C. M., Ellison, J. E., and Garrod, D. R. ((1984). ). Dual role for Dictyostelium contact sites B in phagocytosis and developmental size regulation. Nature 307, , 646–647.[CrossRef][Medline]
Chadwick, C. M., and Garrod, D. R. ((1983). ). Identification of the cohesion molecule, contact sites B, of Dictyostelium discoideum. J. Cell Sci. 60, , 251–266.[Abstract]
Chia, C. P. ((1996). ). A 130-kDa plasma membrane glycoprotein involved in Dict
site, the C-terminal residue of the processed protein, after cleavage and attachment of a GPI-anchor. The cleaved C-terminal signal peptide (29 residues) is shown in bold italics if Asp739 is the 
, DH1; 
, 1C7; 
, 2F5. (E) Colonies of 1C7 and 2F5 after 10 d of growth on lawns of bacteria had larger clear zones than the colonies of DH1. Scale bar, 5 mm.
