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

J. Biol. Chem., Vol. 275, Issue 28, 21331-21339, July 14, 2000

This Article Abstract Full Text (PDF) All Versions of this Article: 275/28/21331    most recent M909642199v1 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 HighWire Citing Articles via Google Scholar Google Scholar Articles by Vitale, N. Articles by Vaughan, M. Search for Related Content PubMed PubMed Citation Articles by Vitale, N. Articles by Vaughan, M. Social Bookmarking                      What's this?

Specific Functional Interaction of Human Cytohesin-1 and ADP-ribosylation Factor Domain Protein (ARD1)^*

Nicolas Vitale^§, Gustavo Pacheco-Rodriguez^¶, Victor J. Ferrans, William Riemenschneider, Joel Moss, and Martha Vaughan

From the Pulmonary-Critical Care Medicine Branch and the  Pathology Section, NHLBI, National Institutes of Health, Bethesda, Maryland 20892

Received for publication, December 6, 1999, and in revised form, February 18, 2000

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Activation of ADP-ribosylation factors (ARFs) is mediated by guanine nucleotide-exchange proteins, which accelerate conversion of inactive ARF-GDP to active ARF-GTP. ARF domain protein (ARD1), a 64-kDa GTPase with a C-terminal ADP-ribosylation factor domain, is localized to lysosomes and the Golgi apparatus. When ARD1 was used as bait to screen a human liver cDNA library using the yeast two-hybrid system, a cDNA for cytohesin-1, a ~50-kDa protein with ARF guanine nucleotide-exchange protein activity, was isolated. In this system, ARD1-GDP interacted well with cytohesin-1 but very poorly with cytohesin-2. In agreement, cytohesin-1, but not cytohesin-2, markedly accelerated [³⁵S]guanosine 5'-3-O-(thio)triphosphate binding to ARD1. The effector region of the ARF domain of ARD1 appeared to be critical for the specific interaction with cytohesin-1. Replacement of single amino acids in the Sec7 domains of cytohesin-1 and -2 showed that residue 30 is critical for specificity. In transfected COS-7 cells, overexpressed ARD1 and cytohesin-1 were partially colocalized, as determined by confocal fluorescence microscopy. It was concluded that cytohesin-1 is likely to be involved in ARD1 activation, consistent with a role for ARD1 in the regulation of vesicular trafficking.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

ARD1,¹ a 64-kDa member of the ADP-ribosylation factor (ARF) family of GTPases, is composed of a 46-kDa GAP domain in the N-terminal region and an 18-kDa ARF domain at the C terminus (1, 2). ARFs have been classified into three groups according to their molecular size and sequence identity (3). All can activate cholera toxin ADP-ribosyltransferase (4) and a specific mammalian phospholipase D (5), and serve as critical components in cellular vesicular trafficking pathways (4). The ARF domain of ARD1 possesses an effector region similar to that of ARFs (6) that contains an amphipathic -helix linked to a -sheet (7, 8) and that is not present in other members of the Ras superfamily. Myristoylation of the N terminus of ARFs seems to play an important part in their membrane association (9). ARD1 lacks a sequence corresponding to the ARF N-terminal -helix and myristoylation site. The structural similarity, however, suggests that ARD1 may have a physiological function comparable to that of ARFs.

Transport between membrane compartments is achieved by vesicular trafficking using COPI-, COPII-, or clathrin-coated vesicles (10-12). Proteins synthesized in the endoplasmic reticulum are transported to the Golgi via COPII-coated vesicles. Intra-Golgi transport and bidirectional traffic between the Golgi system and the endoplasmic reticulum is effected through COPI-coated vesicles (12, 13). Clathrin-coated vesicles are involved in transport among the plasma membrane, endosomes, trans-Golgi compartments, and lysosomes (14). ARF1 participates in the generation of COPI- and clathrin-coated vesicles, whereas Sar1 has an analogous role in COPII vesicle formation. ARD1 is associated with the Golgi apparatus and lysosomes and presumably functions in transport between these organelles (15).

The priming event in vesicle formation involves the activation of a GTPase by a specific guanine nucleotide-exchange protein (GEP). Guanine nucleotide exchange on ARD1 is slowed by a centrally located GDP dissociation-inhibitory domain (16), and the activation of ARD1, like that of ARFs, probably requires its interaction with a GEP. Several ARF GEPs have been identified. They fall into two classes, based on molecular size and sensitivity to inhibition by the fungal fatty acid metabolite brefeldin A (BFA). The family of ~200-kDa BFA-inhibited ARF GEPs includes mammalian BIG1 and BIG2 (17-19) and yeast Gea1/Gea2 (20), Sec7 (21), and GBF1 (22). The BFA-insensitive ARF GEPs are members of the ~50-kDa cytohesin family, which comprises cytohesin-1 (23), cytohesin-2 (also known as ARNO) (24), and cytohesin-3 (also known as GRP1) (25).

All ARF GEPs contain a central, catalytically active Sec7 domain that accelerates guanine nucleotide binding to ARF1. The cytohesins also contain a ~60-residue N-terminal region and a C-terminal pleckstrin homology domain, which binds phospholipids and could have a regulatory function (reviewed in Ref. 26). The Sec7 domains of cytohesin-2 (27, 28), Gea2 (29), and cytohesin-1 (30) have the same overall conformation, with two domains of five -helices each. The active site is located in the C-terminal portion, and residues critical for catalysis are included in sequences termed motifs 1 and 2, which are highly conserved among ARF GEPs. In a three-dimensional structure of the Sec7 domain of Gea2 associated with a truncated form of human ARF1, amino acids in the switch I and II regions of ARF participate in specific interactions with the Sec7 domain (29). Significantly, ARD1 contains all of the critical residues involved in these interactions.

The shared identities of sequence and biochemical properties of ARD1 and ARFs prompted us to search for an ARD1 GEP. Here, we provide evidence that ARD1 efficiently and specifically interacts with cytohesin-1, in vitro and in vivo, as shown by a yeast two-hybrid assay. We further identify the critical residues of ARD1 and cytohesin-1 that are involved in these specific interactions.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Materials-- Polymerase chain reaction (PCR) reagents and restriction enzymes, unless otherwise indicated, were purchased from Roche Molecular Biochemicals. Pfu polymerase was from Stratagene. Sources of other materials have been published (15, 31). Recombinant ARF1 was synthesized and purified as described (32). PcDNA3.1 (ARD1) and pEGFP-C2 (ARD1) constructs have been described (15). Recombinant cytohesin-1, cytohesin-2, cytohesin-1 Sec7 domain, and cytohesin-2 Sec7 domain were synthesized as His₆ fusion proteins (33), and p3 and p3 (24-30 ARF1) were synthesized as GST fusion proteins (6).

Construction and Expression of N-terminal-His ARD1-- To synthesize ARD1 with an N-terminal His₆ tag, ARD1 (GenBank^TM accession number L04510) from pGEX-5G/LIC (2) was amplified by PCR using Pfu polymerase, forward primer 5'-CACTTACCATATGGCTACC CTGATTGTAAACAAGCTC-3' and reverse primer 5'-GAATTCCCGGGGATCCAACTGCG-3'. The forward primer introduced a NdeI restriction site (italicized sequence) with an initiation codon in-frame (boldface sequence), and the reverse primer introduced a BamHI restriction site (italicized sequence). Taq polymerase was introduced at the last cycle, and the PCR fragment was ligated into the pCR^TM vector by TA cloning according to the manufacturer's instructions (Invitrogen). The pCR^TM (ARD1) plasmid was digested with NdeI and BamHI for 1 h at 37 °C; the ARD1 fragment was extracted from LM-agarose gels, purified with a Wizard kit (Promega), and ligated in-frame to the NdeI- and BamHI-digested and dephosphorylated pET-14b expression vector (Novagen). DH5-competent Escherichia coli (Life Technologies, Inc.) were transformed with the plasmid pET-14b (ARD1). The entire sequence of the ARD1 construct was confirmed by automated sequencing (31).

For large-scale production of fusion proteins, 10 ml of an overnight culture of BL21-competent cells (Novagen) containing the pET-14b (ARD1) plasmid were added to 1 liter of LB broth containing ampicillin, 100 µg/ml, followed by incubation at 37 °C with shaking. When the culture reached an A₆₀₀ of 0.6, 500 µl of 1 M isopropyl--D-thiogalactopyranoside were added (final concentration of isopropyl--D-thiogalactopyranoside, 0.5 mM). After incubation for 3 h more, bacteria were collected by centrifugation (Sorvall GSA, 6000 rpm at 4 °C for 10 min) and stored at 20 °C. Bacterial pellets were dispersed in 10 ml of cold phosphate-buffered saline, pH 7.4, containing 20 µg/ml trypsin inhibitor, 5 µg/ml leupeptin, 5 µg/ml aprotinin, and 0.5 mM phenylmethylsulfonyl fluoride. Lysozyme (20 mg in 30 ml) was added, and after 30 min at 4 °C, cells were disrupted by sonification and centrifuged (Sorvall SS34, 16,000 rpm at 4 °C for 20 min). N-terminal-His-ARD1 denatured in 7 M urea was affinity-purified from inclusion bodies on a Ni²⁺ column according to the manufacturer's instructions (Novagen), followed by dialysis for stepwise reduction of the urea concentration (34). Proteins were purified by gel-filtration through Ultrogel AcA54 before storage in small portions at 20 °C. Purity was estimated by Coomassie Blue staining after SDS-polyacrylamide gel electrophoresis and found to be >95%. Amounts of protein were estimated by the Bradford method (35).

Site-directed Mutagenesis-- Point mutations were created in pET-14b (ARD1), pEG202 (ARD1), pQE30(C-1Sec7), and pQE30(C-2 Sec7) with a Quick-Change site-directed mutagenesis kit from Stratagene. Mutations and sequences of the entire clones were confirmed by automated sequencing. ARD1(T418N) and ARD1(K458I) mutations yielded GDP-bound and GTP-bound mutant proteins, respectively. His-ARD1(T418N), His-ARD1(K458I), His-cytohesin-1Sec7(Q4K), His-cytohesin-1Sec7(K30Q), His-cytohesin-1Sec7(Q38R), His-cytohesin-2Sec7(K4Q), His-cytohesin-2Sec7(Q30K), and His-cytohesin-2Sec7(R38Q) proteins were synthesized and purified as described for nonmutant proteins.

Yeast Two-hybrid Screening-- Human ARD1 from pcDNA3.1myc/His(ARD1) (15) was amplified by PCR using Pfu polymerase forward primer 5'-GGGAGACCCAAGCTGGATCCCT ATGGCTACCCTGATTGT-3' and reverse primer 5'-TTCTGAGATGAGTTCTCGAGCTCAA GCAACATCCAATACTCCAGC-3'. The forward primer introduced a BamHI restriction site, and the reverse primer introduced a XhoI site (italicized sequences). The ARD1 PCR fragment was cloned in pEG202 yeast plasmid (Origene Technologies Inc., Rockville, MD). In EGY48 cells, (MAT  trp1his3 ura3 leu2:::6 LexAop-LEU2), ARD1, ARD1(T418N), and ARD1(K458I) did not autoactivate the LEU2 reporter gene or the reporter plasmid pSH18-34 (lacZ) from the DupLEX-ATM yeast two-hybrid system (Origene Technologies Inc.). Expression of ARD1, ARD1(T418N), and ARD1(K458I) in yeast was confirmed by Western blotting. The ability of Lex-A-ARD1, Lex-A-ARD1(T418N), and Lex-A-ARD1(K458I) fusion proteins to enter the nucleus and to bind Lex-A operators was confirmed by a repression assay.

EGY48 cells containing pEG202(ARD1) and pSH18-34 plasmids were transformed further with a human liver cDNA library (Origene Technologies, Inc.). Colonies were selected for growth on His/Leu/TrpUra medium containing galactose. Transformants were further selected for galactose growth dependence and lacZ expression. Plasmid DNA was isolated from yeast by phenol extraction and introduced (Gene Pulser, Bio-Rad) into KC8-competent cells, which were selected on minimal medium plates without Trp. Potentially positive cDNA was introduced into EGY48 cells, and specificity of interaction was tested by mating with RFY206 cells (MATa trp1::hisG his3200 ura3-52 lys2201 leu2-3) containing ARD1-pEG202, ARD1(T418N), ARD1(K458I), or negative control plasmids. cDNAs were sequenced as described above.

Cloning of Cytohesin-2 and Cytohesin-1-- Full-length human cytohesin-2 cDNA was amplified by PCR with Pfu polymerase from a human liver cDNA library (Origene Technologies, Inc.) and ligated into pJG4-5 yeast expression vector. Full-length human cytohesin-1 cDNA was amplified by PCR with Pfu polymerase from the clone cytohesin-1-pJG4-5 and ligated into pcDNA3.1/myc/His(A) expression vector (Invitrogen).

Antibodies-- ARD1 antibodies and secondary antibodies have been described (15, 31). Sec7 antibodies were raised by immunization of a rabbit with recombinant Sec7 domain of cytohesin-1.

Guanine Nucleotide-binding Assay-- Guanine nucleotide binding by ARD1 and ARF1 proteins was assessed by a rapid filtration procedure. Samples were incubated for the indicated time at 37 °C in 20 mM Tris, pH 8.0, 1 mM DTT, 1 mM EDTA, 2 mM MgCl₂ with 0.3 mg/ml ovalbumin, 0.2 mg/ml L--phosphatidyl-L-serine (plus 1 mg/ml cardiolipin, for assays of ARD1) and 4 µM [³⁵S]GTPS or [³H]GDP (~3 × 10⁶ cpm; total volume, 100 µl) with GEP proteins as indicated. Samples were transferred to nitrocellulose filters for rapid filtration, followed by five washes, each with 1 ml of ice-cold buffer (25 mM Tris-HCl, pH 8.0, 100 mM NaCl, 1 mM DTT, 1 mM EDTA, 5 mM MgCl₂). Dried filters were dissolved in scintillation fluid for radioassay.

Assay of GTPase Activity-- Samples were incubated for 30 min at 30 °C in 20 mM Tris, pH 8.0, 10 mM DTT, 2.5 mM EDTA with 0.3 mg/ml bovine serum albumin and 1 mg/ml cardiolipin to promote release of bound nucleotide(s) in preparation for binding of GTP. Then, the samples were incubated for 40 min at 30 °C in the same medium with 0.5 µM [-³²P]GTP (3000 Ci/mmol) and 10 mM MgCl₂ (total volume, 120 µl). After dilution with 380 µl of buffer and incubation for the indicated time at 30 °C, proteins with bound nucleotides were collected on nitrocellulose. Bound nucleotides were eluted in 250 µl of 2 M formic acid, of which 3-µl samples were analyzed by TLC on polyethyleneimine-cellulose plates.

Assay of Cholera Toxin-catalyzed ADP-Ribosylagmatine Formation-- ARD1 proteins were incubated for 30 min at 30 °C in 40 µl of 20 µM Tris, pH 8.0, 10 mM DTT, 2.5 mM EDTA with 0.3 mg/ml bovine serum albumin and 1 mg/ml cardiolipin before addition of 20 µl of a solution to yield final concentrations of 100 µM GTPS and 10 mM MgCl₂. Components needed to quantify ARD1 stimulation of cholera toxin-catalyzed ADP-ribosylagmatine formation were then added in 70 µl to yield final concentrations of 50 mM potassium phosphate (pH 7.5), 6 mM MgCl₂, 20 mM DTT, 0.3 mg/ml ovalbumin, 0.2 mM [adenine-¹⁴C]NAD⁺ (0.05 µCi), 20 mM agmatine, 1 mg/ml cardiolipin, and 100 µM GTPS with 0.5 µg of cholera toxin. After incubation at 30 °C for 1 h, samples (70 µl) were processed as described by Pacheco-Rodriguez et al. (33).

Cell Culture and Expression of Recombinant Proteins-- COS7 cells were grown in Dulbecco's modified Eagle's medium containing 25 mM glucose, 10% fetal bovine serum, penicillin and streptomycin (each at 10 units/ml), and 200 mM glutamine (15). Expression plasmids with DNA encoding cytohesin-1, ARD1, or ARD1 mutant proteins were introduced into COS7 cells (24-well dishes, 80% confluent) using LipofectAMINE Plus (Life Technologies, Inc.), according to the manufacturer's instructions. After incubation for 2-3 h at 37 °C, 1 ml of culture medium with fetal bovine serum and antibiotics was added. Expression of cytohesin-1, ARD1, and ARD1 mutant proteins was assessed after 48 h by immunofluorescence microscopy. Transfection efficiency, estimated for each transfection by counting 500 cells in randomly selected planar sections, ranged from 1 to 15%.

Immunocytochemistry-- Cells grown on coverslips were fixed for 20 min with 4% paraformaldehyde in 0.12 M sodium/potassium phosphate buffer, pH 7.0 (36). After several rinses with PBS, cells were permeabilized for 4 min in PBS containing 0.1% Triton X-100 and then incubated for 1 h with 3% bovine serum albumin and 10% normal goat serum in PBS. Cells were then incubated in a moist chamber for 1 h with the primary antibody diluted in PBS containing 3% bovine serum albumin, washed, and incubated with the appropriate secondary antibodies diluted 1:400. Coverslips were then extensively washed with PBS, rinsed with water, and mounted in Mowiol 4-88 (Hoechst). For evaluation of immunofluorescence, samples were inspected with a PlanApo oil immersion objective (× 60) on a Nikon ELWD0.3 microscope. Acquisition of labeled cell images was accomplished with a Leica laser scanning confocal microscope (15). No staining was observed with the secondary antibody alone or when cells had not been permeabilized or had been mock-transfected. All observations were repeated at least once with a different preparation of cells. Photomicrographs are representative of at least 95% of transfected cells, except as indicated in Fig. 10.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Interaction of Human Cytohesin-1 with Human ARD1-- As a first step to identify proteins that interact with ARD1 in vivo, a human liver cDNA library was screened using Lex-A yeast two-hybrid method (37), with ARD1 as bait. Clones positive for both galactose-dependent LeuG cell growth and -galactosidase expression were isolated. One isolated clone encoded the full-length cytohesin-1 cDNA. No other cDNA of the cytohesin family was identified. To evaluate specificity of the ARD1-cytohesin-1 interaction, cytohesin-1 and cytohesin-2 (ARNO) were compared in the two hybrid system. Both proteins contain the Sec7 and pleckstrin homology domains characteristic of this family of proteins, are overall 83% identical in amino acid sequences, and are similarly effective in accelerating guanine nucleotide binding to class I ARFs (23, 24). EGY48 cells expressing cytohesin-1 or cytohesin-2 were mated with RFY206 cells expressing ARD1 and grown on His/Leu/Trp/Ura medium containing galactose. Diploid colonies were transferred to new plates, and -galactosidase expression was evaluated (Fig. 1A). Cytohesin-1 activated the lacZ gene >60-fold in 20 experiments, whereas cytohesin-2 and two negative controls did not (Fig. 1B); this pattern is consistent with a specific ARD1 cytohesin-1 interaction. An affinity-purified antibody raised against the Sec7 domain of cytohesin-1 (Fig. 1C, lane 5) reacted with both cytohesin-1 and cytohesin-2 (Fig. 1C, lanes 4 and 5) but not detectably with any yeast Sec7-related protein in EGY48 cells (Fig. 1C, lane 1). These data suggest that cytohesin-1 and cytohesin-2 proteins were present at similar levels in the transfected cells (Fig. 1C, lanes 2 and 3).

View larger version (46K): [in this window] [in a new window]   Fig. 1.   ARD1 interacts specifically with cytohesin-1 in a Lex-A-based yeast two-hybrid assay. A, RFY206 cells expressing ARD1 were mated with EGY48 cells transformed with pJG4-5, pJG4-5 cytohesin-1, pJG4-5-cytohesin-2, or pRFM1. Diploid cells were grown on HisG and LeuG medium and selected for galactose growth dependence before lacZ expression was estimated. B, results of 20 independent experiments were quantified by densitometry (Personal Densitometer SI, Molecular Dynamics). Data are means ± S.E. of values expressed relative to that of the strong activator plasmid pSH17-4 100%. C, equal amounts (~20 µg of protein) of yeast EGY48 cells (lane 1, nontransfected; lane 2, expressing cytohesin-1; lane 3, expressing cytohesin-2) were suspended in Laemmli buffer, boiled for 10 min, and subjected to SDS-polyacrylamide gel electrophoresis in 4-20% gels. Lanes 4-6 contain 1 µg of recombinant cytohesin-1, cytohesin-2, or cytohesin-1 Sec7 domain protein, respectively. Proteins were transferred to nitrocellulose membranes, which were incubated with an antibody against cytohesin-1 Sec7 domain. Positions of protein standards (in kDa) are indicated on the right.

Biochemical Characterization of the Interaction between ARD1 and Cytohesin-1-- Cytohesin-1 and cytohesin-2 accelerated GTPS binding to ARF1 to the same extent (Fig. 2A), whereas cytohesin-1 was much more effective than cytohesin-2 in stimulating GTPS binding to ARD1 (Fig. 2B), consistent with the specificity of activation of ARD1 by cytohesin-1. Cytohesin-1 increased GTPS binding to ARF1 much more than to ARD1. A ~200-kDa BFA-sensitive ARF GEP (17) was also a relatively poor activator of GTPS binding to ARD1 relative to ARF1 (data not shown). The high concentrations of the two GEPs (equimolar with potential substrates) enabled us to detect a small effect of cytohesin-2 on GTPS binding by ARD1.

View larger version (22K): [in this window] [in a new window]   Fig. 2.   Effect of cytohesin-1 and cytohesin-2 on GTPS binding by ARF1 and ARD1. 20 pmol of recombinant ARF1 (A) or ARD1 (B) were incubated without (open circles) or with 1 µg (~21 pmol) of cytohesin-1 (closed circles) or cytohesin-2 (closed squares) at 37 °C with 4 µM [35S]GTPS for the indicated time. Data are means ± S.E. of values from triplicate assays. The experiment was repeated three times with similar results.

Mutant forms of ARD1, which were expected to have dominant negative or positive effects, were generated by site-specific mutation. The ARD1(T418N) and ARD1(K458I) mutants are GDP-bound and GTP-bound mutant proteins, respectively. As expected, the binding of GTP and GDP to ARD1(K458I) was very similar, whereas GTP binding to ARD1(T418N) was minimal (Fig. 3). ARD1 possesses intrinsic GTPase activity as a consequence of its GAP domain (31), and ARD1(T418N) maintained the ability to hydrolyze GTP (despite poor GTP binding), whereas ARD1(K458I) had no detectable GTPase activity (data not shown). ARD1, like ARFs, is a GTP-dependent activator of the ADP-ribosyltransferase activity of CTA (1). In the presence of GTP, ARD1(K458I) stimulated the formation of ADP-ribosylagmatine slightly more than did ARD1, probably because it did not hydrolyze the bound GTP. As expected, ARD1(T418N) did not significantly activate CTA, presumably as a result of its inability to bind GTP significantly (Fig. 3). In the presence of GDP, all mutants were relatively inactive.

View larger version (18K): [in this window] [in a new window]   Fig. 3.   Characterization of ARD1-GDP and ARD1-GTP mutants. 50 pmol of recombinant ARD1, ARD1(T418N), or ARD1(K458I) were incubated with 4 µM [35S]GTPS or [3H]GDP for the indicated time, and nucleotide binding was determined as in Fig. 2. Insets show activation of CTA ADP-ribosyltransferase by the ARD proteins. 80 pmol of ARD1, ARD1(T418N), or ARD1(458I) were incubated with 100 µM GTP or GDP for 20 min at 30 °C. Components needed to quantify ARF stimulation of CTA-catalyzed ADP-ribosylagmatine formation were then added, and incubation was continued for 1 h at 30 °C. CTA activity in the absence of ARD1 has been subtracted. Data are means of values from quadruplicate assays ± S.E.

The abilities of ARD1 mutants to interact with cytohesin-1 and cytohesin-2 were evaluated in a Lex-A-based yeast-two hybrid system. The interaction of cytohesin-1 appeared considerably greater with the GDP-bound mutant ARD1 and was essentially zero with the GTP-bound mutant (Fig. 4). In vitro cytohesin-1 stimulated binding of GDP (Fig. 5B), but not of GTPS (Fig. 5A), by ARD1(T418N), whereas binding of both GTPS and GDP by ARD1(K458I) was increased (Fig. 5). These results suggest that in cells, cytohesin-1 interacts preferentially with the GDP-bound form of ARD1, and these results are consistent with the current view that GEPs stabilize a nucleotide-free form of GTP-binding proteins (38).

View larger version (28K): [in this window] [in a new window]   Fig. 4.   Cytohesin-1 interacts preferentially with the ARD1-GDP mutant in a Lex-A-based yeast two-hybrid assay. RFY206 cells transfected with ARD1, ARD1(T418N), ARD1(K458I), or pRFMH1 constructs were mated with EGY48 cells expressing cytohesin-1 or cytohesin-2. Diploid cells were grown on HisG and LeuG medium and selected for galactose growth dependence. Data are means ± S.E. of values for lacZ expression in 20 independent experiments analyzed as described in Fig. 1.

View larger version (22K): [in this window] [in a new window]   Fig. 5.   Effect of cytohesin-1 on nucleotide binding by ARD1 mutants. Twenty pmol of ARD1 (circles), ARD1(T418N) (squares), or ARD1(K458I) (triangles) were incubated without (open symbols) or with (closed symbols) 1 µg (~21 pmol) of cytohesin-1 at 37 °C with 4 µM [35S]GTPS (A) or [3H]GDP (B) for the indicated times. Data are means ± S.E. of values from triplicate assays. The experiment was repeated twice with similar results.

Regions of ARD1 Involved in the Specific Interaction with Cytohesin-1-- Physical interactions with cytohesin-1 were evaluated using GST fusion proteins of the separate ARD1 domains. No interaction of cytohesin-1 or cytohesin-2 with GST (Fig. 6) or GST-p5 (GAP domain of ARD1) was detected (data not shown). GST-p3, the ARD1 domain that corresponds to an ARF lacking 15 N-terminal amino acids, interacted specifically with cytohesin-1 but not with cytohesin-2 (Fig. 6), confirming results of the initial yeast two-hybrid screen and in vitro assays. Amino acids 24-30 in the ARF domain of ARD1 are very different from those in the equivalent positions (39-45) of ARF1, which are highly conserved among the ARFs (6). This sequence is located in the so-called effector region of other Ras superfamily members, in which it interacts with regulatory proteins. A GST fusion protein of a chimeric protein termed p3(24-30ARF1), in which residues 24-30 in the ARF domain of ARD1 were replaced with the corresponding sequence of ARF1, interacted with both cytohesin-1 and cytohesin-2, but to a lesser extent than did the ARF domain of ARD1 with cytohesin-1 (Fig. 6). These results could reflect a contribution of the ⁴²⁶QDEFMQP⁴³² sequence in ARD1 to its specific interaction with cytohesin-1.

View larger version (18K): [in this window] [in a new window]   Fig. 6.   Binding of cytohesin-1 and cytohesin-2 to immobilized ARF domain of ARD1. GSH-Sepharose beads with 10 µg of bound GST, GST-p3 (ARF domain of ARD1), or GST-p3(24-30ARF1) (in which amino acids 24-30 of p3 were replaced by the corresponding ARF1 sequence) were incubated with 10 µg of cytohesin-1, cytohesin-2, or bovine serum albumin (BSA) in 500 µl of buffer containing 20 mM Tris (pH 8), 1 mM DTT, 20 mM NaCl, 1 mM MgCl2, and 2 mM EDTA for 1 h at 37 °C. Beads were washed four times with 1 ml of buffer and eluted with 100 µl of Laemmli buffer, of which 20-µl samples were subjected to SDS-polyacrylamide gel electrophoresis in a 4-20% gels followed by staining with Coomassie Blue. Positions of protein standards (in kDa) are indicated on the left (MW). Findings have been reproduced twice with two different preparations of proteins. *, densitometry showed that binding of cytohesin-1 and cytohesin-2 to GST-p3(24-30ARF1) was only ~60% of that of cytohesin-1 to GST-p3.

Possible Mechanism of Guanine Nucleotide Exchange-- As reported, ARF lacking the N-terminal amphipathic -helix can form a stable complex with the Sec7 domain of cytohesin-1 (18) or cytohesin-2 (39). Complex formation between p3 and C-1Sec7 appeared to be influenced by the nucleotide bound to the GTPase. Physical association of the ARF domain of ARD1 with cytohesin-1 was highly dependent on experimental conditions. The presence of 1 mM GTP or GDP, or a millimolar concentration of free Mg²⁺, completely prevented physical association of the two proteins (data not shown). Thus, a free Mg²⁺ concentration in the micromolar range, which favors nucleotide release from ARD1 and accumulation of the nucleotide-free form, also favored the formation of a complex between ARD1 p3 and the cytohesin-1 Sec7 domain.

Amino Acids Involved in Cytohesin-1 Interaction with ARD1-- In vitro, the Sec7 domain of cytohesin-1 can accelerate nucleotide binding to human ARF1, ARF5, and ARF6 (33, 39). The Sec7 domains of cytohesin-1 (C-1Sec7) and cytohesin-2 (C-2Sec7) were equally effective in stimulating GTPS binding to ARF1 and other ARFs (data not shown). C-1Sec7, like the intact protein, stimulated GTPS binding to ARD1, whereas cytohesin-2 or its Sec7 domain did not (data not shown). Accordingly, C-1Sec7 interacted physically with GST-p3, whereas C-2Sec7 did not (Fig. 7). On the other hand, both C-1Sec7 and C-2Sec7 interacted with the chimeric protein GST-p3 (24-30ARF1), although to a lesser extent than did C-1Sec7 with GST-p3 (Fig. 7), consistent with the presence of a region responsible for the specificity of interaction with ARD1 within the cytohesin-1 Sec7 domain.

View larger version (39K): [in this window] [in a new window]   Fig. 7.   Interaction of cytohesin-1 and cytohesin-2 Sec7 domains with the ARF domain of ARD1. GSH-Sepharose beads with 10 µg of bound GST, GST-p3 (ARF domain of ARD1), or GST-p3(24-30ARF1) were incubated with 10 µg of C-1Sec7 or C-2Sec7 before interacting proteins were separated and stained as described in Fig. 6. Identical results were obtained with two different protein preparations.

The most striking structural difference between the effector region of the ARF domain of ARD1 and those of the other ARFs is the presence of a negatively charged residue (Asp-427) in ARD1 that corresponds to Gly-40 in ARF1. The preferential association of ARD1 with cytohesin-1 could result from the interaction of Asp427 with a positively charged residue present in the Sec7 domain of cytohesin-1, but not cytohesin-2. Conversely, cytohesins that do not interact with ARD1 may contain in that position a residue that cannot pair with Asp-427. The Sec7 domains of the known members of the cytohesin family share 74% identity and 85% similarity of amino acid sequences (Fig. 8A). In positions 4, 30, and 38 of the Sec7 domain (Fig. 8A, boxed), cytohesin-1 differs in charge from the other cytohesins. We replaced singly each of these amino acids in C-1Sec7 with the corresponding residue in cytohesin-2 and made the converse mutations in C-2Sec7. All mutant proteins similarly accelerated GTPS binding to ARF1 (data not shown). Each mutant protein was tested for interaction with GST-p3. The mutant C-2Sec7(Q30K) interacted with GST-p3 (Fig. 8B), whereas no interaction was detected with C-1Sec7(K30Q), C-2Sec7, C-2Sec7(K4Q), or C-2Sec7(R38Q). Mutations in positions 4 and 38 did not prevent the interaction of C-1Sec7 with GST-p3. C-1Sec7(K30Q), however, was inactive, and C-2Sec7(Q30K) effectively stimulated GTPS binding to ARD1 (Fig. 8C). Activities of the other mutant proteins were not different from those of the corresponding nonmutated proteins (Fig. 8C). These results strongly support the idea that Lys-30 in the Sec7 domain of cytohesin-1 is important in its interaction with ARD1, perhaps through an ionic interaction with Asp-427, although the evidence for its involvement is less compelling than that for Lys-30. Position 30 in the cytohesin Sec7 domains is in the loop between helices A and B, and it seems possible that association of C-1Sec7 with ARD1 results from a conformational change following initial interaction of these two proteins.

View larger version (76K): [in this window] [in a new window]   Fig. 8.   Identification of amino acid in the cytohesin-1 Sec7 domain responsible for specific interaction with ARD1. A, alignment of amino acid sequences of Sec7 domains of cytohesin-1 (C-1), cytohesin-2 (C-2), cytohesin-3 (C-3), and cytohesin-4 (C-4). The three positions in which residues differ in charge are boxed. B, GSH-Sepharose beads with 10 µg of bound GST-p3 were incubated with 10 µg of C-1Sec7 (C-1 WT) C-2Sec7(C-2 WT), C-1Sec7(Q4K) (4, C-1), C-2Sec7(K4Q) (4, C-2), C-1Sec7(K30Q) (30, C-1), C-2Sec7(Q30K) (30, C-2), C-1 Sec7(Q38K) (38, C-1), and C-2Sec7(R38Q) (38, C-2). Interacting proteins were separated and stained as described in Fig. 6. Similar results were obtained twice. C, recombinant ARD1 (20 pmol) was incubated alone (×); with 1 µg (~45 pmol) of C-1Sec7 (), C-1Sec7(Q4K) (), C-1Sec7(K30Q) (), or C-1Sec7(Q38R) (); or with 1 µg of C-2Sec7 (), C-2Sec7(K4Q) (), C-2Sec7(Q30K) (), or C-2Sec7(R38Q) () at 37 °C with 4 µM [35S]GTPS for the indicated times. Data are means ± S.E. of values from triplicate assays. The experiment was repeated twice with similar results.

Intracellular Interaction of Cytohesin-1 with ARD1-- As reported (15), ARD1 overexpressed in COS7 cells appeared to be associated with vesicular structures, some of which were concentrated around the nucleus and others of which were scattered throughout the cytoplasm, apparently corresponding to the Golgi apparatus and lysosomes, respectively (Fig. 9, A-C) (15). ARD1(T418N), which preferentially binds GDP, had a distribution very similar to that of ARD1 (Fig. 9, D-F), whereas ARD1(K458I), the GTP-bound form, was concentrated close to the nucleus (Fig. 9, G-I). Overexpressed cytohesin-1 was more widely distributed, seemingly associated with some structural elements, and was also found in the nucleus (Fig. 9, J-L).

View larger version (109K): [in this window] [in a new window]   Fig. 9.   Intracellular distribution of ARD1, ARD1 mutants, and cytohesin-1. COS7 cells transfected with pEGFP-ARD1 (A-C), pEGFP-ARD1(T418N) (D-F), pEGFP-ARD1(K458I) (G-I), or pcDNA3.1myc-His-cytohesin-1 (J-L) are shown. Cytohesin-1 is red, and ARD1 is green. Nomarski images are shown alone (B, E, H, and K) or superimposed on the immunofluorescence images (C, F, I, and L). Similar results were obtained with at least four different COS7 cell preparations and with NIH 3T3 cells. Bar, 20 µm.

When both ARD1 and cytohesin-1 were overexpressed in the same cells (Fig. 10, A-D), ARD1 was apparently concentrated in the perinuclear region and lysosomes and colocalized with cytohesin-1 to only a very limited extent (Fig. 10D). Cytohesin-1 was distributed throughout the cells in pattern resembling that seen when it was expressed without ARD1 (Fig. 9, J-L). The distribution of ARD1(T418N) co-expressed with cytohesin-1 was dramatically different from that of ARD1. In ~90% of cells, it was widely distributed, excepting the nucleus (Fig. 10E), and appeared largely co-localized with cytohesin-1 (Fig. 10H). In fewer than 10% of cells expressing both proteins, ARD1(T418N) was concentrated in presumably lysosomal structures (Fig. 10I) and to a large extent co-localized with cytohesin-1 (Fig. 10J). Thus, ARD1(T418N) and cytohesin-1 seemed to be absent from Golgi but associated with lysosomal structures (Fig. 10, E-L). ARD1(K458I) was mainly concentrated around the nucleus consistent with a Golgi localization (Fig. 10M), as it was when expressed alone (Fig. 9G). Co-expressed cytohesin-1 (Fig. 10, M-P) was in Golgi-like elements but was also spread throughout the cell in a pattern similar to that seen when it was expressed alone (Fig. 9J). These results are consistent with the conclusion that cytohesin-1 and ARD1 can interact when overexpressed in COS7 cells, as well as in yeast. In many cells, co-expression of GFP-ARD1(T418N) and myc-His-cytohesin-1 altered the distribution of both proteins. It was notable that the number of cells containing both proteins was much less than the number expressing only one or the other.

View larger version (115K): [in this window] [in a new window]   Fig. 10.   Intracellular distribution of ARD1 and ARD1 mutants in cells co-expressing cytohesin-1. COS7 cells expressing cytohesin-1 (A-P) and ARD1 (A-D), ARD1(T418N) (E-L), or ARD1(K458I) (M-P) are shown. Cytohesin-1 is red, and ARD1 is green. Nomarski images are shown alone (C, G, K, and O) or with the immunofluorescence images superimposed (D, H, L, and P). Yellow results from colocalization. E-H and I-L are representative of 90 and 10%, respectively, of cells expressing cytohesin-1 and ARD1-GDP. Similar results were obtained with at least six different cell preparations and with two different plasmid preparations. Bar, 20 µm.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

In an attempt to find proteins that interact with ARD1, we identified cytohesin-1, an ARF GEP, as a potential regulator of ARD1 activation. Specific interaction was suggested by the yeast two-hybrid screen, in which cytohesin-2 failed to interact like cytohesin-1 with ARD1. Biochemical characterization identified certain regions of both cytohesin-1 and ARD1 that define specificity for their molecular contacts. Cytohesin-1 was initially characterized as a GEP by its ability to increase guanine nucleotide binding to ARFs of class I, namely ARF1 and ARF3, and also ARD1 (23, 33). It has been difficult, however, to associate this activity of cytohesin-1 with a specific intracellular function. Although effects of cytohesin-1 on the interaction of LFA1 with its ligand ICAM1 (40) were reported before its GEP activity was recognized (23), the relationship between the two remains to be elucidated. Ashery et al. (41) recently reported that cytohesin-1 in murine brain participates in functions related to ARF6, although other workers have failed to show that cytohesin-1 efficiently accelerates guanine nucleotide binding to ARF6 in vitro (33, 42). Thus, the physiological functions of this GEP (and others) are still unclear.

Several mammalian ARF GEPs have been cloned, and the proteins have been characterized. These include BFA-insensitive cytohesin-1, cytohesin-2 (ARNO), cytohesin-3 (ARNO3 and GRP1), and EFA6 (43) and BFA-inhibited BIG1, BIG2, and GBF1, but additional information is needed regarding the tissue distribution, intracellular localization, and regulatory interactions. Attempts to demonstrate subcellular localization of individual ARFs and GEPs have resulted in differences of opinion, for example, regarding the role of cytohesin-2 (ARNO) as a GEP for ARF6 (44, 45). This could be the result of differences in assay conditions. However, all of the known ARF GEPs appear to act on ARFs of class I, specifically ARF1 or ARF3. Although GEP activity resides in the Sec7 domain, at least some determinants of substrate recognition clearly exist outside of that region (33). It appears that identification of specific ARF/GEP pairs solely by in vitro assays will not be possible until the effects of intracellular conditions on the ARF-GEP interaction are understood much better than they are now.

ARD1, originally cloned by screening a cDNA library with an ARF2 cDNA (1), turned out to be a 64-kDa protein with a C-terminal 18-kDa ARF domain. Later, studies identified part of the remaining 46-kDa as a GAP for the ARF domain (2), which contains a zinc finger motif (31). Interaction between the ARF and GAP domains is mediated in part by the ARF effector region (6). Association of a GAP with the corresponding effector or switch region of a substrate is not unexpected (46). ARD1 residues ⁴²⁶QDEFMQP⁴³², which confer specificity for cytohesin-1, differ distinctly from the corresponding ARF sequences. Furthermore, it has been shown that lysine 38 in ARF1 is critical for ARF1 interactions with either the full-length cytohesin-1 or its Sec7 domain (47). This residue, which corresponds to lysine-425 ARD1, is present in all of the ARF proteins thus far identified as substrates for the Sec7 domain of cytohesin-1. The finding that both the GAP domain of ARD1 and cytohesin-1 interact with the same ARF region suggests that additional structural elements play regulatory roles in these two interactions. For example, the GDP dissociation inhibitory region of ARD1 (6) may contribute to this interaction, which seems likely because the equivalent region in ARF1 is required for its interaction with cytohesin-1 (33). The demonstration that a mutant of ARD1 containing only the ARF domain plus the 15 amino acids at its N terminus was a substrate for the intact cytohesin-1 (33) is consistent with this view and with the conclusion that multiple regions of the ARD1 molecule are involved in the interaction with cytohesin-1 Sec7.

Both cytohesin-1 and its Sec7 domain increased guanine nucleotide binding to full-length ARD1 (33). By contrast, cytohesin-2 and its Sec7 domain interacted very poorly. Although the Sec7 domains are highly conserved among cytohesins, there are major amino acid differences at positions 65, 92, and 100 (numbers in cytohesin-1). These residues are situated within the first five -helices and appear to account for a large part of the functional differences between cytohesin-1 and cytohesin-2. They do not seem to be critical for activity, because their replacement in cytohesin-1 with those of cytohesin-2 did not alter activity with ARF1 as a substrate. Thus, it appears that the motifs 1 and 2 of the Sec7 domain are critical for activity, and the N terminus contributes to specificity. The x-ray structure of the Sec7 domain of yeast Gea2, a BFA-inhibited ARF GEP (20), co-crystallized with a truncated form of human ARF1, did not reveal extensive interactions of the first five -helices with ARF1 (48). We found that ARD1(T418N) (ARD1-GDP), but not ARD1(K458I) (ARD1-GTP), associated specifically with cytohesin-1, and all of the data are consistent with the notion that the interaction occurs preferentially with nucleotide-free ARD1, which appears to be true also for GEP interactions of other members of the Ras GTPase superfamily.

ARD1, endogenous and overexpressed, had been localized to lysosomes, and to a lesser extent to Golgi membranes, by fluorescence microscopy and subcellular fractionation (15). In contrast, recombinant cytohesin-1 was distributed more diffusely throughout the cells, perhaps associated with some kind of cytoskeletal structures, but usually not with the homogeneous pattern expected of a cytosolic protein. Colocalization of the two proteins in cells over expressing both was limited and was seen clearly only when ARD1(T418N), which is believed to exist with GDP bound, was used (Fig. 10). The subcellular distribution of cytohesin-1 described here, and its apparent concentration just under the plasma membrane in some cells (data not shown), may reflect its reported interaction with <2 integrin and effects on cell adhesion (40). How this is related to the postulated role of ARD1 in vesicular transport or trafficking between Golgi and lysosomes (15) remains to be determined.

	ACKNOWLEDGEMENTS

We thank Ronald Adamik for antibodies against the cytohesin-1 Sec7 domain, Dr. Elisabetta Meacci for the cytohesin-2 clone, and Carol Kosh for expert secretarial assistance.

	FOOTNOTES

^* The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

These authors contributed equally to this work.

^§ Present address: INSERM U-338 Biologie de la Communication Cellulaire, 5 rue Blaise Pascal, 67084 Strasbourg Cedex, France.

^¶ To whom correspondence should be addressed: Rm. 5N-307, Bldg. 10, 10 Center Dr., MSC 1434, National Institutes of Health, Bethesda, MD 20892-1434. Tel.: 301-402-1454; Fax: 301-402-1610; E-mail: PachecoG@nih.gov.

Published, JBC Papers in Press, March 16, 2000, DOI 10.1074/jbc.M909642199

	ABBREVIATIONS

The abbreviations used are: ARD1, ARF domain protein; ARF, ADP-ribosylation factor; BFA, brefeldin A; C-1Sec7, cytohesin-1 Sec7 domain; C-2Sec7, cytohesin-2 Sec7 domain; CTA, cholera toxin A subunit; DTT, dithiothreitol; GEP, guanine nucleotide-exchange protein; GST, glutathione S-transferase; GTPS, guanosine 5'-3-O-(thio)triphosphate; COPI, coat protein complex I; COPII, coat protomer II; PCR, polymerase chain reaction.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				A. Vichi, D. M. Payne, G. Pacheco-Rodriguez, J. Moss, and M. Vaughan E3 ubiquitin ligase activity of the trifunctional ARD1 (ADP-ribosylation factor domain protein 1) PNAS, February 8, 2005; 102(6): 1945 - 1950. [Abstract] [Full Text] [PDF]

				A. Someya, M. Sata, K. Takeda, G. Pacheco-Rodriguez, V. J. Ferrans, J. Moss, and M. Vaughan ARF-GEP100, a guanine nucleotide-exchange protein for ADP-ribosylation factor 6 PNAS, February 27, 2001; 98(5): 2413 - 2418. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 275/28/21331    most recent M909642199v1 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