Identification of amino acid residues at the interface of a ...

2 downloads 3 Views 2MB Size Report
Searle Mark I1 Model 6847 scintillation counter. Isolation of the regA ...... Adam, S. A., Naka awa, T Swanson, M. S., Woodruff, T. K., and Dreyfuss,. 6. Dreyfuss, G.

Vol. 267, No. 36, Issue of December 25, pp. 26097-26103. 1992

THEJOURNAL OF BIOLOGICAL CHEMISTRY 0 1992 by The American Society for Biochemistry and Molecular Biology, Inc

Printed in U.S.A.

Identification of Amino Acid Residuesat the Interface of a Bacteriophage T4 regA Protein-Nucleic Acid Complex* (Received for publication, August 19, 1992)

Kevin R. WebsterS, Sarah Keill, William Konigsberg, Kenneth R. Williams$, and Eleanor K. Spicery From the Department of Molecular Biophysics and Biochemistry and the $Howard Hughes Medical Institute, Yale University School of Medicine, New Haven, Connecticut 0651 1

The bacteriophageT4 regA protein (M, = 14,600) is a translationalrepressor of a group of T4early mRNAs. To identify a domain of regA protein that is involved in nucleic acid binding, ultraviolet light was used to photochemically cross-link regA proteinto [s2P]p(dT)16. Thecross-linked complex was subsequently digested with trypsin, and peptides were purified using anion exchange high performance liquid chromatography. Two tryptic peptides cross-linked to [sap]p(dT)le were isolated. Gas-phase sequencing of the major cross-linked peptide yielded the following sequence: VISXKQKHEWK, which corresponds to residues 103-113 of regA protein. Phenylalanine 106 was identified as the siteof cross-linking, thusplacing this residue at the interface of the regA protein-p(dT),,, complex. The minor cross-linked peptide corresponded to residues 31-41, and the site of cross-linking in the peptide was tentativelyassigned to Cys-36. The nucleic acid binding domain of regA protein was further examined by chemical cleavage of regA protein into six peptides using CNBr. Peptide CN6, which extends from residue 95 to 122, retains both the ability to be cross-linked to [32P]p(dT)16 and 70% of the nonspecific binding energyof the intact protein. However, peptide CN6 does not exhibit the binding specificity of the intact protein. Three of the other individual CNBr peptides have no measurable affinity for nucleic acid, as assayed by photo-cross-linking or gel mobility shifts.

not previously been examined is the domain structure of regA protein and the aminoacid residue(s) involved in RNA binding. A number of RNA binding proteins have been shown to contain common structural motifs. For example, a family of eucaryotic RNA binding proteins sharea common structural feature termed the RNP’ domain (or RNA recognition motif, RMM), which consists of approximately 80 amino acids including4 well conserved phenylalanine residues (5-10). Within the 80-aminoacid domain is an octapeptidesequence of basic and hydrophobicresidues,includingaconserved phenylalanine,thatispostulatedto play a role inRNA recognition. Thisdomainisfoundinsmall nuclear R N P proteins, hnRNP proteins,poly(A) binding proteins, and nucleolin (7). Asecond type of RNAbindingdomainthathas been identified is characterized by an arginine-rich motif which is common t o a number of procaryotic and eucaryotic RNA binding proteins (11).These proteins include the bacteriophage N proteins, HIV tat and rev proteins, retroviral gag proteins, and someribosomal proteins (11).The arginine-rich domain has been implicated in the specific recognition of RNA stem-loop structures by the N protein during bacteriophage transcription antitermination events and in HIV tat protein-RNA interactions(11).In fact,a 14-residue synthetic peptide which contains the arginine-rich sequences of tat protein has beenshown to bind ina specific fashion to a target RNA stem-loop structure (the tar element)(l2). Thus, it is now apparent that both the ability to bind nucleic acids and to exhibit specificity of binding, in somecases, may require only a small protein fragment. T h e bacteriophage T 4 regA protein regulates the expression regA protein does not exhibit significant similarity with own synthesis, of at least 12 T4 genes while also regulating its either of the RNA binding domains described above. The fact at the level of translation (1).It has been shown that regA that regA protein recognizes a specific group of mRNAs protein acts by binding to specific mRNAs and competing (unlike R N P domain proteins) and that its target RNAelewith the formationof initiation complexes by ribosomes (2). ment is single-stranded (unlike the elements recognized by N Furthermore, it is known that regA protein acts by the recprotein, tat andrev) suggests that a different structuralmotif ognition of a specific RNA element, inwhich apparently both may be involved in RNA recognition and binding. To localize the sequence and structure areof importance (3,4). What has the RNA binding domain of regA protein,photochemical *This research was supported byArmy Research Office Grant cross-linking hasbeen used to identify amino acid residues at DAAL03-91-G-0293(to E. K. S.)and by National Institutes of Health the interface of a regA protein-nucleic acid complex. In adGrants GM12607 (to W. H. K), GM30191 (to E. K. S.),and GM37573 dition, thenucleic acidbinding characteristicsof regA protein (to K. R. W.). This project was also supported in part by Biomedical fragments were examined to further defineregions of the Research Support Group Grant RR 05358 (to E. K. S.) awarded by protein likely to be involved in RNA binding. ~

~

~

~~~

the Division of Research Resources, National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This articlemust therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. $ Current address: Bristol-Myers Squibb Co., Pharmaceutical Research and Development Division, Dept. of Cellular & Molecular Biology, 5 Research Parkway, P.O. Box 5101,Wallingford, CT 06492. ll To whom correspondence and reprint requests should be addressed. Tel.: 203-737-2567;Fax: 203-737-2638.

MATERIALS AND METHODS

Protein Purification-regA protein ( M , = 14, 600) was purified from AR120 cells containing plasmid pASlregA following induction of transcription from the lambda P L promoter by nalidixic acid The abbreviations used are: RNP, ribonucleoprotein; HIV, human immunodeficiency virus; HPLC, high performance liquid chromatography; PTH, phenylthiohydantoin; CAM, carboxamidomethyl.

26097

26098

regA Protein-Nucleic Acid Interactions

treatment (13), as previously described (14). The protein was stored in 20 mM Tris/HCl (pH 7.5),150 mM NaC1, 5 mM MgClz, 1 mM EDTA, 1 mM dithiothreitol, and 50% glycerol a t a concentration of 0.7-1.1 mg/ml, at -70 "C. Concentration of regA protein beyond 1.4 mg/ml resulted in precipitation of the protein. One g of nalidixic acid-induced cells yielded -10 mg of pure regA protein. ) irradiated Photochemical Cross-linking-regA protein (4 p ~ was in 25 pl of buffer A (10 mM in the presence of 2.8 p M [32P]p(dT)16 Tris, 75 mM NaCl, 2.5 mM MgC12,0.5 mM dithiothreitol, 5% glycerol, pH 7.5). Samples were irradiated with a germicidal UV lamp (X -253 nm) at a height of 5 cm (generating 5 X lo3 ergs/mm*/min) at 4 "C ina Petri dish. The efficiency of cross-linking as a function of exposure time was quantitated using a filter binding assay (15). Whatman GF/c41 filter paper was presoaked in cold 15% trichloroacetic acid for 30 s and allowed to dry for 2 h before use. Irradiated samples were applied to the dried filters and allowed to air-dry for 5 min. Filters were then rinsed in a Buchner funnel under vacuum with 50 ml of cold 10% trichloroacetic acid, 50 ml of cold dH20 (3 times), and 20mlof cold 100% ethanol (three times). Filters were subsequently dried under a heat lamp for 15 min and transferred to vials containing 7 ml of scintillation fluid. The samples were counted in a Searle Mark I1 Model 6847 scintillation counter. Isolation of the regA Tryptic Peptide-P2P]p(dT),, Cross-linkedComplex-regA protein (315 nmol) was mixed with 200nmol of [3zP] P ( ~ T )(specific ~ G activity = 20,800 cpm (Cerenkov)/nmol) in 6 ml of buffer A and allowed to stand for 10 min. at 4 "C. The sample was then irradiated for 20 min a t 4 "C, with a resultant UV dose of 1.0 X 10' ergs/mm2. The extent of cross-linking and recovery of complex during purification were quantitated by liquid scintillation counting (Cerenkov counts). The cross-linking reaction was transferred to a centrifuge tube andthen precipitated by addition of cold 100% trichloroacetic acid to give a final concentration of 10% (v/v). The sample was then centrifuged and the pellet was washed twice with cold acetone. The pellet was resuspended in 1.2 ml of 8 M urea, 50 mM NH4HC03 andsonicated in a water bath sonicator for 30 min. To allow for identification of cysteine during sequencing, the peptide was carboxamidomethylated by first incubating with 120 plof 45 mM dithiothreitol for 15 min a t 50 "C and thenadding 120 pl of 100 mM iodoacetamide and incubating for 15 min at 25 "C (16). The reaction was then diluted to reduce the concentration of urea to 2 M. A 1:25 (w/w) ratio of tosylamide-2-phenylethylchloromethylketone-treated trypsin (Cooper-Biomedica1):cross-linkedprotein was added, and the sample was incubated at 37 'C for 24 h. The tryptic peptide mixture was then injected ontoa Nucleogen DEAE 60-7 anion exchange column, equilibrated in buffer B (20 mM NaAc, 20% CH&N (pH 6.5)). regA protein tryptic peptides were eluted at a flow rate of 1ml/ min with a linear gradient from 0 to 1.5 M KC1 in buffer B over 55 min. Fractions containing potential cross-linked peptides were desalted using Sep-Pak CIScartridges (Waters Associates) equilibrated in 5 mM triethylammonium acetate (pH 7.0) and eluted with 5 mM triethylammonium acetate (pH 7.0):50% methanol. Fractions were stored at 4 "C. Amino Acid Analysis and Protein Sequencing-Amino acid composition and concentrations of peptide and protein samples were determined by amino acid analysis by the Keck Foundation Biotechnology Resource Laboratory at Yale University. Duplicate samples of 1-5pgwere dried and then dissolved in 0.1 ml of 6 N HCl, 0.2% phenol, with 1nmol of norleucine added as an internalstandard. The sample was then incubated at 150 "C for 1.5h. The hydrolyzed material was dried, resuspended in 0.1 ml of citrate buffer (sodium diluent 2.20 Pickering), and applied to a Beckman 7300 amino acid analyzer. For protein sequencing, samples eluted from Sep-Pak Cls cartridges in 5 mM triethylammonium acetate (pH 6.8) were concentrated -4fold by Speed Vac. A portion of the sample was submitted to amino acid analysis to determine peptide concentration, and a second portion was loaded onto a cartridge filter precycled with Biobrene (Applied Biosystems) and sequenced using an Applied Biosystems Model 470A gas-phase Protein Sequencer. CNBr Cleavage-A preparative reaction with CNBr was performed using 9 mg (615 nmol) of regA protein. Prior to cleavage, the protein was trichloroacetic acid-precipitated as described above, and the dry protein was resuspended in 1.8 ml of 88%formic acid. A 10001 molar ratio of CNBr to methionine residues was used for cleavage: 500 pl of 5 M CNBr(in CH3CN) was added tothe protein, reducing the concentration of formic acid to 70%. Reaction with CNBr was performed at 25 "C for 24 h in the dark. The sample was subsequently diluted 4-fold, dried to a volume of approximately 100 pl, and then

diluted to a volume of 1 ml with 0.1% trifluoroacetic acid, 6 M guanidine HCl. The sample was then boiled for 3 min and injected onto a CISreverse-phase column equilibrated in 0.1% trifluoroacetic acid. Peptides were eluted with a linear gradient of 0-60% CH3CN in 0.1% trifluoroacetic acid at a flow rate of 1ml/min over 60 min. The resultingpeaks were identified by amino acid composition and aminoterminal sequencing. Peptides were stored at 4 "C. Fluorescence Spectroscopy-regA protein fluorescence was detected on an SLM Model 8000 spectrofluorometer, interfaced with an IBM PC-XT computer. Titrations were performed in Buffer C (10 mM Hepes (pH 7.2), 5 mM MgC12,l mM EDTA, 1mM P-mercaptoethanol) plus a specified concentration of NaCl unless otherwise noted. Data were acquired at theexperimentally determined excitation maximum of 282 nm and emission maximum of 347 nm. The effects of photobleaching throughout the titration were corrected for by monitoring the peptide fluorescence in a control cuvette. Screening of incident light by poly- and oligonucleotides was corrected for by a parallel titration of N-acetyl-L-tryptophanamide(Sigma). All data were acquired through "reverse" titrations (the addition of lattice to ligand) (17). Theapparent association constant (Kapp)was calculated as follows. regA protein] [regA-lattice complex] = (%Q/%QQmax)[total %Q = ( F , - F,)/F; X 100

K. = [regA-lattice complex]/[free sites][free regA]

(Eq. 1)

The binding site size for peptide CN6 was assumed to be -9 as has been determined for regA protein (3). Although it might be predicted that peptide CN6 would have a somewhat smaller occluded site size than regA, the limiting amount of this peptide that was available and its relatively weak affinity for nucleic acids precluded a direct determination of its site size. However, evenif the actual site size for CN6 is only 5-4, this would result in only an -2-fold error inthe calculated affinities, which is similar to our estimatederror in determining replicate constants. All data were examined using a double reciprocal analysis in which l/AF was plotted versus l/[free oligo] (18,19). The Y intercept of the plot yields l/AFmaX for the particular titration, and the relationship also allows the determination of K,

K,, = (m .AFmax)"

(Eq. 2)

where m = slope. Gel Retardation A~say-5-~~P-labeledgene 44 RNA 16-mer (5AAUGAGGAAAUUAUCA-3)(0.5 p ~ was ) incubated with regA protein (1p ~ or) a CNBr fragment (1-10 p M ) in 20 mM Hepes (7.2), 10 mM NaCl, 5 mM MgCl,, 16% glycerol, for 5 min at 25 "c.Samples were transferred to ice for 5 min and then electrophoresed on a nondenaturing 4% acrylamide:88 mM Tris borate (pH 8.3) gel. RNA fragment migration was visualized by autoradiography. RESULTS

Photochemical Cross-linking of regA Protein to PzPl p(dT)16-Complexes of regA protein and aradiolabeled target RNA termed g44-4 (5'-AAUGAGGAAAUU-3'), containing the gene 44 recognition element for regA protein (3), were exposed to increasing doses ofUV light. Formation of covalently cross-linked complexes wasthen monitored by electrophoresis on SDS-polyacrylamide gels. Although cross-linking was observed, the efficiency of cross-linking (

Suggest Documents