Peptide Analogues Compete with the Binding of &-Factor to Its [PDF]

Vol. 263, No. 33, Issue of November 25, pp. 17333-17341,1988 Printed in U.S.A.

THEJOURNAL OF BIOLOGICAL CHEMISTRY 0 1988 by The American Society for Biochemistry and Molecular Biology, Inc.

Peptide Analogues Compete with theBinding of &-Factor toIts Receptor inSaccharomyces cereuisiae* (Received for publication, April 13, 1988)

Susan K. RathsSQ,Fred Naiderll, and Jeffrey M. BeckerSII From the $Department of Microbiology and the 11 Program in Cellular, Molecular, and Developmental Biology, University of Tennessee, Knoxville, Tennessee 37996 and the WDepartment of Chemistry, College of Staten Island, City University of New York,Staten Island, New York 10301

a-Factor, a secreted tridecapeptide pheromone, is required for mating between the a-and a-haploid mating typesof Saccharomyces cerevisiae. An analogue of a-factor, [DHPS,DHP",Nle12]tridecapeptide (where DHP represents 3,4-dehydro-~-proline and Nle represents norleucine), wascatalytically reduced inthe presence of'H gas to produce a radiolabeled pheromone with high specific activity, purity, and biological activity. Association and dissociation kineticsindicated values of 4.9 X lo" M" s-' for kl and 1.1 X lo-' s-' for k-'. Saturation binding studies gave equiliban rium dissociation constant equalto 2.3 X lo-' M, which approximated the kinetically derived KO of 2.2 X M. These values compare favorably to the previously determined KO of 6 X lo-' M (Jenness, D. D., Burkholder, A. C., and Hartwell, L. H. (1986) Mol. Cell. dissociation Biol. 6, 318-320). Scatchard analysis and in the presence of excess unlabeled ligand indicated interaction with a homogeneous population of noninteracting binding sites (13,000 sites/cell). A number of a-factor analogues, previously investigated for their structure-function relationships (Naider, F., and Becker, J. M. (1986) CRC Crit. Rev. Biochem. 21, 225-249),were used to compete with ['Hla-factor binding. Fourtridecapeptideshavingconservative amino acid replacements bound strongly to thereceptor. In contrast,[Phe'la-factor and 10 des-Trp'-a-factor analogues bound tothereceptor 1-3 orders of magnitude less effectively than did a-factor itself. The binding constants for all active pheromones correlated with biological activity. However, des-Trp'[Phe']afactor anddes-Trpl-[Alas]a-factor, which were not biologically active, stillcompeted with a-factor binding, indicating that these analogues fail to induce a secondary signal necessary for biological response tothe pheromone. One analogue, des-Trp'-[Cha',~-Ala*]afactor (where Cha represents cyclohexylalanine), was not biologically active and did not demonstrate binding to the receptor, whereasdes-Trp'-[Cha',~-Ala~]a-factor was active and bound to the receptor. This finding suggests that a type I1 &turn is necessary for binding of a-factor toits receptor and for subsequent biological activity.

* This work was supported in part by Grants GM-22086 and GM22087 from the National Institute of General Medical Sciences. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. f Predoctoral fellow supported by National Institutes of Health Training Grant AI-07123.

The yeast Saccharomyces cereuisine is a eukaryote which can exist either in a haploid or diploid state and canundergo both asexual and sexual reproduction. The reciprocal action of two diffusible peptide pheromones, a- and a-factors, mediate the events necessary for mating by the two haploid cell types a and a. When exposed to the appropriate pheromone, each haploid exhibits a variety of characteristic responses (1) including cell-surface agglutinin synthesis, cell-cycle arrest in GI,and morphological transformation termed the shmoo in preparation for cell and nuclear fusion (2-4). a-Factor, the tridecapeptide pheromone synthesized constitutively by acells and active on a-cells, was originally isolated from a-cell culture supernatants by Duntze e t al. (5) and later found to be encoded by twogenes, MFal andMFa2 (6, 7). The sequence encoded by these genes, WHWLQLKPGQPMY, was chemically synthesized and found to exhibit all activities attributed to thebiological pheromone (8, 9). Jenness et al. (10) showed that partially purified, biologically produced %-labeled a-factor bound specifically to acells and not to a-cells or ala-diploids. This analysis also revealed that, among a number of temperature-sensitive sterile mutants, only the ste2 mutant showed binding activity which was stable at permissive temperature and unstable at restrictive temperatures. Using chemically synthesized and tritiated a-factor, Jenness e t al. (11)reported an equilibrium dissociation constant of 6 X lo-' M and 8000 binding sites for a-factorlcell. The STE2 gene wascloned by complementation in two laboratories (12, 13), and the sequence was found to predict an integral membrane protein with seven transmembrane-spanning regions. That this gene product is the ligandbinding component of thea-factor receptor was recently established by both biochemical and genetic techniques (15, 16). Thus, the a-factor receptor appears similar to a family of receptors which includes the &adrenergic and muscarinic acetylcholine receptors and rhodopsin (14). In this study, we continue our effort to understand the primary interaction of a-factor with its receptor. Our previous studies (17) as well as those from the laboratory of Masui et al. (18) have evaluated consequences of amino acid changes in the a-factor solely by their effect on biological activities, i.e. agglutination, growth arrest, and shmoo induction (for a review, see Ref. 17). In this report, we present binding constants for a variety of a-factor analogues and relate these constants tobiological activity of the pheromones. Our results represent the first example of an equilibrium binding study of the relative affinities of various a-factor analogues and provide insights into the molecular interactions and stereochemical requirements which influence binding to the a-factor receptor.

17333

17334

Binding of S. cerevisiae a-Factor Analogues MATERIALS AND METHODS'

Organisms, Media, ana' Growth Conditions

S. cerevisiae strains 4202-15-3 (MATa cryl barl-1 ade2-l his4-580 lys2 trpl tyrl SUP4-3) and 4202-1-2 (MATa cryl barl-I- ade2-1 his4-580 lys2 trpl SUP4-3) used in this study were obtained from Duane Jenness (Department of Molecular Genetics and Microbiology, University of Massachusetts Medical School, Worchester, MA). Strains were maintained on (YM-1) agar containing, yeast growth medium in grams/liter: yeast extract, 5; peptone, 10; yeast nitrogen base without amino acids (filter-sterilized), 6.7; adenine, 0.01; uracil, 0.01; succinic acid, 10; sodium hydroxide, 6; glucose, 10; agar, 20, at a final pH of 5.8 (19). Cells used in the binding assays were grown in liquid YM-1 at 30 "C and 200 rpm overnight in a rotary water bath (New Brunswick Scientific Co., Edison, NJ) to a density no greater than 1 X lo7 cells/ml (A550 = 0.4). Inhibitor medium (YM-l+i) was YM-1 with the addition of 10 mM NaN3, 10 mM KF, and 10 mM ptosyl-L-arginine methyl ester as previously described (10). Inhibitor medium wasroutinely prepared and filtered through Millipore filters (type HA, 0.45 p) immediately before use.

peptides containing Nle in place of Met were obtained in 25% yield based on starting resin. In our laboratory, we demonstrated complete reduction of the 2 dehydroproline residues in [DHP8,DHP",Nle'2]tridecapeptideusing hydrogen gas and palladium black in methanol in 2 h as judged using 400-MHz NMR and monitoring resonances associated with the double bond of DHP (data not shown). The reduced tridecapeptide was isolated by HPLCand shown to be identical to [Nle'2]a-factor. Further discussion on the biological and chemical properties of the radioactive peptide is included in the Miniprint. The radioactive tridecapeptide was prepared from [DHPa,DHP1',Nle'']tridecapeptide by catalytic tritiation by Amersham Corp. by the TR3 method. The crude tritiated product was purified to homogeneity as described in the Miniprint. a-Factor Biological Activity Assays

Halo Assay for Growth Arrest-S. cerevisiae 4202-15-3or 4202-1-2 cells (1 X lo6) were plated using 4 ml of 0.8% top agar for each minimal medium plate (2% Noble agar; 2% glucose;0.67% yeast nitrogen base without amino acids; 20 pg/ml each adenine, histidine, and tryptophan; and 30 pg/ml lysine and tyrosine). Whatman No. 1 filter paper discs containing the specific amounts of the indicated a-Factor Peptides peptides were placed on each plate. After incubation for 36 h at 30 "C, a-Factor is a tridecapeptide of the following structure: H2N-TrpHis-Trp-Leu-Gln-Leu-Lys-Pro-Gly-Gln-Pro-Met-Tyr-COOH. Ana- halos of growth arrest were measured as described (21). Shmoo Assay-S. cerevisiae 4202-15-3or 4202-1-2cells were grown logues of this peptide are designated according to IUPAC convention to slightly less than 1 X lo7 cells/ml in YM-1 overnight in a rotary (20). Thus, if norleucine (Nle)' is substituted for methionine at water bath at 30°C. Cells were harvested by centrifugation ina position 12, the analogue is designated [Nle'2]tridecapeptide or [Nle"] Beckman Model TJ-6 table-top centrifuge (1500 X g, 5 min), washed a-factor. The removal of tryptophan at residue 1 gives des-Trp'-aonce with fresh medium, and resuspended to a final assay concentrafactor. Other residues used for substitution were cyclohexylalanine tion of 2 X lo6 cells/ml. Costar 96-well microtiter plates were used to (Cha), 5-dimethylaminonaphthalene-1-sulfonyl (DNS), and 3,4-de- make 2-fold dilutions of each peptide in YM-1. The final concentrahydro-L-proline (DHP). tion range tested for each a-factor peptide was 25 pg/ml to 0.15 ng/ All dodecapeptide a-factor analogues used in this study were syn- ml. Once 100 p1 of the cell stock (4 X lo6 cells/ml) were added to the thesized by solution-phase techniques as described previously (21100 pl of peptide solution in each well, the plates were incubated at 25). These peptides were greater than 98% homogeneous on reversed- 30 "C on a rotary agitator. At 4.5 h, samples were removed from each phase HPLC, gave one ninhydrin-positive, UV-positive spot on silica well for microscopic examination. Affected cells were identified to thin layers, and gave the expected amino acid ratios. The tridecapepcomparison to control wells containing cells without a-factor peptide. tides used in this study were a-factor, [Nle'2]tridecapeptide, [Phe3] The activity end point was defined as the last well showing cells at tridecapeptide, [Asn6,Arg7]tridecapeptide,and [DHPa,DHP",Nle'2] 4.5 h with a large pointed projection from the cell body typically tridecapeptide. Allwere prepared using the solid-phase method of called the shmoo morphology (29). synthesis. In these syntheses, the first amino acid was attached to a standard Merrifield (chloromethyl polystyrene, 1%divinylbenzene) PHla-Factor Binding Assay or a (pheny1acetamido)methyl resin. The chain was extended using Binding of [3H]a-factorto 4202-15-3 or 4202-1-2 cells was assayed either dicyclohexylcarbodiimide or dicyclohexylcarbodiimide l-hydroxybenzotriazole as the coupling agent with t-butoxycarbonyl pro- using the basic protocol of Jenness et al. (10)with a few modifications. tection on the a-amine. Deprotection was accomplished using 40% YM-l+i (the medium used in binding studies) contained NaN3, KF, trifluoroacetic acid in methylene chloride. For peptides containing and p-tosyl-L-arginine methyl ester (see "Materials and Methods"), methionine, 2% dimethyl sulfide (v/v) was added to the acidolysis which effectively stops energy metabolism preventing a-factor intersolvent. The resulting trifluoroacetate was neutralized using 10% nalization but not interfering with a-factor binding (10,30). Thecelldiisopropylethylamine in methylene chloride. Coupling reactions were mediated degradation of a-factor has been discussed by numerous monitored using qualitative ninhydrin analysis. Usually complete authors (31, 32). In this present study, the strain used were bar1 coupling was attained in the first step. Nevertheless, we routinely (sstl), which lacks the peptidase that cleaves the mating factor (32); double-coupled all residues. The overall procedure follows that de- and protease and metabolic inhibitors were included in the binding scribed previously (26). The following side chain protecting groups assays. During binding assays under conditions used in this study, no were utilized in these syntheses: tosyl for histidine, 2-chlorobenzylox- hydrolysis of [3H]a-factor was seen (data not shown). We previously ycarbonyl for lysine, and 2,6-dichlorobenzylor 2-bromobenzyloxycar- showed that all analogues used in this investigation are degraded bony1 for tyrosine. In certain syntheses, Trp was protected by the more slowly than the native tridecapeptide (17). We conclude thereformyl group. In others, it was left unprotected. Cleavage of the final fore that, under the experimental conditions used in both the bioassay peptide from the resin was accomplished using HF/anisol (9O:lO) at and the binding experiments, degradation and internalization of the 0 "C for 1 h or the low/high HF method (27). Crude peptides were pheromones arenot significant. Specific experimental details are purified by preparative reversed-phase HPLC on a C18 column. All given below. final materials were greater than 98% homogeneous on gradient Time Course for a-Factor Binding HPLC andgave one spot on silica thin layers. A representative HPLC has been published (28). of a-factor and [Asn5,Arg7]tridecapeptide To determine the time course of association and dissociation for a Purified peptides had the anticipated 400-MHz 'H NMR spectrum kinetically derived KD,500 ml of YM-1 in 1-liter Erlenmeyer flasks and gave amino acid ratios within 10% of the theoretical values. The were prewarmed for 3 h at 30 "C. These flasks were inoculated with overall yields for peptides containing Met were15-20%, whereas 4202-15-3 and 4202-1-2 cells (final cell concentrations = 1.5 X lo5 cells/ml) from YM-1 plates for growth overnight at 30 "C in a rotary ' Portions of this paper (including part of "Materials and Methods" water bath. Readings (A.550)and hemocytometer counts of the culture and Figs. 7-9) are presented in miniprint at the end of this paper. indicated when harvest was appropriate. Under these conditions, the Miniprint is easily read with the aid of a standard magnifying glass. generation time was 120 min. Cells were harvested a t 1 X lo7 cells/ Full size photocopies are included in the microfilm edition of the ml by centrifugation at 6000 X g and 4 "C (JA-10 rotor, Beckman Model J-21C centrifuge).The resultant pellets were washed two times Journal that is available from Waverly Press. ' The abbreviations used are: Nle, norleucine; Cha, cyclohexylala- with ice-cold filtered YM-l+i andresuspended to 1.11 X lo9cells/ml nine; DHP, 3,4-dehydro-~-proline; DNS, 5-dimethylaminonaphthal- for a cell stock suspension. The reaction was initiated by the addition M, 9 Ci/mmol) to 1.8 ml of labeled a-factor (7.5 X ene-1-sulfonyl; HPLC, high performance liquid chromatography. All of200plof amino acids and peptides are designated according to IUPAC conven- the cell stock solution. A t 1, 2, 3, 4, 5, 8, 10, 15, and 20 min, 50 pl of the reaction mixture were diluted into 1.5 ml of ice-cold YM-l+i and tion (20).

Binding of S. cerevisiae a-Factor Analogues filtered over 1%bovine serum albumin-presoaked GN-6 Metricel filters (Gelman Sciences, Inc., Ann Arbor, MI). The dilution tube was then rinsed twice with 1.5 ml of ice-cold YM-l+i, with each rinse filtered over the same filter. Finally, 1.5 ml of ice-cold YM-l+i were used to wash the filter directly once. The binding of a-factor tofilters in the absence of cells was less than 100 cpm. Over the course of the reaction, the temperature of incubation was kept constant at 22 "C. On the average, tubes were vortexed once every 5 min. At 20 min, four 10-pl aliquots of the reaction mixture were taken to determine total ligand concentration (total amount bound plus total amount free). To determine the amount of nonspecific binding, a-cells were incubated identically to a-cells. At 20 min, triplicate aliquots were withdrawn, diluted into 1.5 ml of ice-cold YM-l+i, and processed as described above for a-cells. Specific binding was defined as those counts bound to a-cells minus those counts bound to a-cells. In all experiments, the counta/minute bound to a-cells were less than 10% of the counts/minute bound to a-cells. At 23 min, the dissociation phase was initiated by diluting separate aliquots of the reaction mixture 200-fold into (a) temperature-equilibrated YM-l+i only or ( b ) temperature-equilibrated YM-l+i containing a 10-fold molar excess of unlabeled a-factor. At 2.5,5, 7.5, 10,15, 20,30,45,60,90, 120, and 135 min, aliquots were filtered and washed as described previously to determine the amount of a-factor remaining bound. Filters were counted in 5 ml of Bray's scintillant (Research Products International Corp., Mt. Prospect, IL).All assays and manipulations were done in silicanized borosilicate tubes or vials.

Steady-state Saturation Binding To obtain steady-state saturation binding for Scatchard analysis, cells were grown and harvested as described above with the following changes. The cell stock suspension was made 3.75 X 10' cells/ml for a final assay cell concentration of 3 X 10' cells/ml. Dilutions of tritium-labeled a-factor (9 Ci/mmol) were made in order to achieve assay concentrations of 7 X 3 X lo-', 2 X 1.5 X lo-', and 4 X lo-' M [3H]a-factor. Initiationof binding by the addition of label was staggered by 5 min such that each concentration could be processed at 35 min with ample time for the manipulations involved. At 5 min, post-bindinginitiation, two 50-p1 aliquots were taken to determine total a-factor concentration (bound plus free) for each experimental point. Incubation was at 22 "C with intermittent vortexing on the average of once every 5 min. At 35 min, two 150-pl aliquots were taken, diluted, filtered on bovine serum albumin-presoaked GN-6 Metricel filters, and washed as described previously. To determine the level of nonspecific binding, a-cells were tested at the same concentrations using the same method. The totalcounts/minute bound to a-cells were less than 10% of the totalcount/minute bound to a-cells.

17335

Duplicate 150-pl aliquots were taken at 35 min, diluted, filtered, and treated as previously described to determine the amount of a-factor bound. To determine total ligand concentration, two 20-pl aliquots were taken at 5 min for counting. RESULTS

Biological Activities of Unlabeled a-Factor and Its Ana1ogues"Biological activities of a-factor and itsanalogues were determined using two assays (Table I). In general, the results with tridecapeptide analogues using S. cerevisiue 4202-15-3 agree with previously published data using other S. cerevisiue strains (17). Des-Trp'-[Ala3]cr-factor,des-Trp'-[Phe3]a-factor, des-Trp'-[Cha3,~-Leu6]a-factor, anddes-Trp'-[Cha3,~Alag]a-factor are not active for growth arrest or shmoo formation for any S. cerevisiae strains examined to date. The relative specific activities of the other dodecapeptide analogues are the same in various strains. The amount of tridecapeptide causing shmoo formation in our hands (1.2 ng/ ml or 6.7 x 10"O M) with strain 4202-15-3 is about 20 times less than the amount Moore (29) reported for the half-maximal concentration for projection formation in S. cerevisiae strain 2180. The difference in concentrations found by us and Moore could reflect minor differences in assigning abberant cell or the fact that Moore used S. cerevisiae 381G in her studies. Association of PHla-Factor with a-Cells-The time course for association of [3H]a-factorwith a-cells is shown in Fig. 1. Association is complete by 10 min at 22 "C.This experiment was designed so that less than 10%of the [3H]a-factor present was bound. Thus, the data inFig. 1 can be plotted according to thepseudo first-order rate equation to yield kl = 5.9 X lo4 " 1 s-l . Alternatively, if no assumptions aremade concerning the relative concentration of ligand and binding sites, these TABLEI Biological activity of a-factor and itsanalogues Activity Peptide

Growth arrest" 1 la

cm

Competition for Tritium-labeled a-Factor Binding by Unlabeled a-Factor Analogues

Tridecapeptide [AsnS,Arg7]Tridecapeptide [Nle"]Tridecapeptide

Shmoob

10rg

nM

3.6 2.7 1.7 2.7 3.6 1.7 3.6 2.7 1.7 Competition of bound [3H]a-factorby unlabeled a-factor analogues 1.7 was measured by the following protocol. In general, a- and a-cells [DHP8,DHP",Nle'2]Tridecapeptide 3.6 2.6 3.3 2.3 3.4 were grown and harvested as described above for the time course of [Phe3]Tridecapeptide 2.4 1.4 550 a-factor binding. Changes included resuspending the cells in YM-l+i Des-Trp'-dodecapeptide 0 0 >35,000 to 1.25 X 10' cells/ml with a final assay concentration of 1 X log Des-Trp'-[Ala3]dodecapeptide 0 0 >35,000 cells/ml. Final assay concentrations of the unlabeled competitors Des-Trp'-[Phe3]dodecapeptide 2.11.0 1,100 1X 3 x Des-Trp'-[Cha3]dodecapeptide were 3 X lo-', 6 X lo-', 1 X lo-', 3 X lo-', 6 X Des-Trp'-[Cha3,~-Leu6]dodecapeptide 0 0 >35,000 6 X 1 X lo-', 3 X 6 X lo-', and 1 X lo-' M. The 1.0 2,200 or AZz0 Des-Trp'-[Cha3,acetyl-Lys7]dodecapep- 0 concentrations of competitors were determined from the A ~ w tide for each peptide. We calibrated extinction coefficients according to 1.6 0.8 1,100 quantitative amino acid analysis of each competitor solution. For Des-Trp'-[Cha3,0ctanoyl-Lys7]dodecapeptide those analogues containing 2 tryptophan residues, an ezW of 13,000 0 >35,000 was determined. This compares to a value of em = 12,100 calculated Des-Trp'-[Cha3,~-Alag]dodecapeptide 0 1,100 by Jenness et al. (10) and em = 1.9 & 0.3 X lo' calculated by Moore Des-Trp'-[Cha3,~-Alag]dodecapeptide1.7 0.9 Des-Trp'-[DNS-His2,Cha3]dodecapep2.0 0.9 139 (29). For analogues containing no tryptophan residues, an ezzO of 1.9 tide X lo' was determined. A value of em = 5,400 was found for a-factor analogues with 1 tryptophan residue. Typically, the reaction was Growth arrest as measured by the halo assay. For each peptide, started by the addition of 50 pl of [3H]a-factor (6 X M) to 400pl 1,5,8, and10 pg were spotted on discs which were placed upon YMof the cell suspension (1.25 X log cells/ml) and either 25 or 50 pl of 1plates with top agar containing strain 4202-15-3 (1 X lo6 cells). At the appropriate unlabeled a-factor analogue. When necessary, prior 36 h, the zone of growth inhibition was determined by subtracting to the addition of the label, volumes were adjusted to total 450 p1 by the disc diameter from the halo diameter in centimeters. The halo the addition of YM-l+i. Positive control reactions consisted of either sizes for 1 and 10 pg of each peptide are shown. a- or a-cells and labeled a-factor, but no unlabeled competitor. a* Values represent activity end points for shmoo formation in the Cells were tested at competitor concentrations of 1 X lo-', 1 X microtiter assay as described under "Materials and Methods." Serial 1X and 1 X M to determine the level of nonspecific binding. dilutions of each peptide were tested for shmoo-inducing effect. The All incubations were done at 22 "C. Starts of the reactions were end points shown indicate the last well in which morphological effect staggered by 5 min to allow time for the manipulations at 35 min. of the peptide was seen microscopically.

Binding of S. cerevisiae a-Factor Analogues

17336

data can be plotted according to the integrated second-order negligible. This was accomplished using two methods: 1) rate equation to yield kl = 3.0 X lo4 M-' s-'. When fresh a- dilution of the binding reaction at equilibrium by 200-fold so cells were added to the supernatant solution remaining after that any rebinding of radioligand, once dissociated, was not a binding experiment, up to 80% of the total counts were detected and 2) addition of unlabeled competing ligand to the bound to a-cells. Given the fact that, at high cell concentra- dilution buffer so that any rebinding that occurred would tions used in the binding assay, filtering was slowand quench- likelyinvolve the unlabeled ligand (34). Fig. 2 shows the ing was difficult to assess rigorously, we assumed that 100% results for a dissociation time course using both the "infinite" of thetritium counts were associated with a-factor. The dilution method (method 1) and infinite dilution in the presMiniprint detailsthe procedures used to characterize the [3H]ence of excess unlabeled a-factor (method2). When the data a-factor used. replotted as in the inset, virtually in Fig. 2 are transformed and The association constant (kl) can also be determined using identical straight lines are traced. The straight line indicates the half-time for association (tl/z).When tl12 was estimated that a-factor dissociation is a first-order reaction. The k-l from a regression line of the initial association rateand determined from this plot is 7.7 X s-'. Alternatively, substituted into the equation (33) when the half-time for dissociation is obtained from a linear regression of the initial dissociation rate in Fig. 2, a value of s-' is obtained. The average rate constants for 1.4 X dissociation determined by these two methods for two differwhere t1I2 = time at which binding is one-half equilibrium ent experiments are listed in the second column of Table 11. The kinetics observed for association and dissociation are binding, [L] = concentration of free ligand at equilibrium, [R] consistent with a model in which [3H]a-factor-receptor com= concentration of free receptor at equilibrium (total number of receptors minus number of receptors occupied at equilib- plexes are formed via a simple bimolecular reaction. The = concentration of receptor-ligand complexes equilibrium dissociation constant ( K D )therefore can be derium), and [RL] rived from the kinetically determined 121 and k-l. The KD (2.2 at equilibrium, kl was found to equal 5.3 X lo4 M" s-'. The first column of Table I1 lists the average association X lo-' M) inTable I1 is the average KD calculated from rate constants (kl) calculated according to the three methods combinations of kl and k-l in the first and second columns. Steady-state Saturation Binding-Since a given number of discussed above for five independent determinations. Averaging these values, the rate constant for [3H]a-factor associ- cells are expected to possess a finite number of receptors, saturability was assessed by examining binding as a function ation (kl) is approximately 4.9 X io4 X M-' s-'. of increasing radioligand concentration (Fig. 3). Fig. 3 (inset) Time Coursefor Dissociation of PHla-Factor from a-CelLsshows a Scatchard plot derived from the saturation binding In order to determine a dissociation rate for the a-factor/ receptor interaction, experimental conditions were fixed so data. The straight line of the Scatchard plot indicates that that the reassociation of the ligand with the receptor was the interaction of a-factor with its receptor can be characterized by one equilibrium dissociation constant (KO).The slope of the line in the Fig. 3 (inset) yields a KD of 2.3 X lo-' M. Furthermore, the x intercept reveals that a-cells contain approximately 13,000 binding sites/cell. Competition for Binding of PHla-Factor by Unlabeled aFactor and Its Analogues-Competition for [3H]a-fa~tor bind$ 1500 ing was determined by incubating a-cells with a constant (6 X lo-' M ) in the presence of concentration of [3H]a-fa~tor 9W 1000 increasing concentrations of unlabeled a-factor peptides (Figs. 4-6). The affinities of these competitors for the a-factor "' 5004 receptor are reflected in the ICs0 values (the concentration of competitor which effectivelycompetes for 50% of the specific 0 10 20 30 4 0 50 60 ligand binding) and theKOvalues listed in Table111. Of those TIME (MIN.) analogues tested, only des-Trp1-[Cha3,~-Alag]dodecapeptide, FIG. 1. Time course for association of ['Hla-factor with acells. For experimental details, see "Materials andMethods." Specific a biologically inactive analogue, did not compete for binding to the a-factorreceptor. All the other peptides tested, includbinding equals the counts/minute bound to a-cells minus the counts/ ing the biologically inactive des-Trp'-[Ala3]peptide, desminute bound to a-cells. ,n

TABLE I1 Kinetically derived rate constants for PHla-factor association ( k l ) and dissociation (k-J and the eauilibrium dissociation constant f K n ) ~

k, ~

L"

s-I

6.7 X lo4 k 0.7 (pseudo 1st order)b 3.4 X lo4 f 0.9 (2nd order)d 4.6 X lo4 + 0.6 ( t 1 J e ( n = 5)'

~

~

_

_

_

_

_

_

_

_

_

_

_

_

~

k-, ~

K"* ~

S-1

M

7.7 X 10" + 0.2 (1st order)' 1.4 X 10-3 f 0.8 ( n = 2)

2.2 X lo-' +. 0.6 ( n = 6)

This value represents the average KDcalculated according to the equation K D= k-,/kl and using all the possible combinations of k, and k-l listed. Calculated according to thepseudo first-order rate equation. e Calculated according to the first-order rate equation. Calculated according to the second-order rate equation. e Calculated according to themeasured half-time ( tkh) for association or dissociation. n equals the number of experimental determinations. Values listed include standard errors of the mean.

'

Analogues ae a-Factor Binding of S. cereuisic

a

0

20

40

120

60

80 100 TIME (MIN.)

140

160

FIG.2. Time course for dissociation of ['Hla-factor from acells using both infinite dilution (A) and addition of excess unlabeled a-factor (A).The inset shows the derivation of the rate constant for [3H]a-factordissociation using first-order kinetics.

rn

Y 2 W

5 4-

2

32-

2

I -

J

m

L

d o

t

o

* * b * a ~

IO FREE

J - t t n t a * * l

I6

a FACTOR

*

1

*

~

16

'

1

~a

~

~~

~

IO

(MOLAR)

FIG. 3. Saturation binding of ['Hla-factor to a-cells. The inset represents a Scatchard plotof the binding data.KD = 2.3 X lo-' M;n = 13,000 sites/cell. For experimental details, see "Materials and

Methods." Trp1[Phe3]peptide, and des-Trp1-[Cha3,~-Leu6]peptide, competed to varying degrees (KO= 1.3 X to 6.5 X M). The natural sequence and several tridecapeptides showing biological activity equal to the naturalsequence were the best competitors. All curves for competing peptides were found to be parallel using pair-wise t tests ( p = 0.01) and of normal steepness showing 10-90% competition over approximately an 81-fold range of competitor. Once the equilibrium constants were determined, the Gibbs standard free energies were calculated (35) using the standard thermodynamic equation: AG" = -RT In KO, where R = gas constant (1.99 cal/mol degrees), T = temperature (kelvin), and KD = equilibrium dissociation constant ("l) (Table 111). DISCUSSION

In order to examine the binding of a-factor analogues for structure-function comparisons, a highly purified, radiolabeled a-factor is required. Previous studies have utilized biologically produced 35S-labeled a-factorpreparations which were partially purified (10, 31) or a [3H]a-factor preparation direct from the manufacturer (11).In a paper published after submission of our work, Blumer et al. (15) reported using 35Slabeled a-factor in a membrane binding assay. For our study, we prepared alabeled pheromone by subjecting the chemically synthesized precursor peptide [DHP8,DHP",Nle12]tridecapeptide to catalytic reduction in the presence of tritium gas.

17337

This resulted in a mixture of radiolabeled products from which [3H]a-factor of high specific activity (9 Ci/mmol) and purity (>98%) was obtained by HPLC (see Miniprint). The association of [3H]a-fa~tor with the a-factor receptor appeared to be a simple reversible bimolecular reaction with an average kl of 4.9 X lo4 M" s" as determined from pseudo first-order kinetics, the integrated second-order rate equation, and t1,2for association. The kinetics for a-factor dissociation were consistent with the interaction of a single receptor population possessing a fixed affinity for [3H]a-factor. Whether the infinite dilution method was used or excess unlabeled a-factor was present, similar dissociation kinetics were observed giving an average k-1 of 1.1 x s-'. From association and dissociation rate data, a kinetic KD for the binding of [3H]a-fa~tor was calculated as 2.2 X lo-' M. The KO was also determined under conditions in which 10% of the radiolabeled a-factor would be bound maximally, thus fulfilling the primary requirement for Scatchard analysis. The slope of the plot (Fig. 3, inset) allowed the estimation of KD = 2.3 X IO-' M, and the x intercept gave the number of binding sites/cell as 13,000 sites/cell. Replotting the regression line according to the suggestion of Klotz (37) generates a binding isotherm that mirrors the experimentally obtained curve (data not shown). As expected for reversible binding of a radioligand to a single class of receptors, the equilibrium dissociation constant determined from steady-statesaturation binding data is equivalent to theKO determined from the kinetic data. Using similar methodology and the same cells for binding studies, but a different labeled a-factor, Jenness et al. (11) estimated a KDof 5-7 X lo-' M and 8,000 receptor sites/cell, and Chvatchko et al. (31) calculated a KD of 1 X lo-' M and 8,000 sites/cell. In previous investigations (36) on structureactivity relationships on the effect of a-factor on cellular ~ ~ J agglutinability and morphogenesis, we concluded that more than one receptor mediated response to thepheromone. The binding studies reported herein as well as by others (10, 11, 31) provide direct evidence concerning the number and type of receptors present in S. cereuisiue. Together with recent genetic (16) and biochemical (15) studies on the STE2 gene, they provide convincing support for the existence of only one a-factor receptor. To quantitate the potency of unlabeled a-factor analogues in competing for [3H]a-factor binding, the ICs0 values (the concentration of competitor which effectivelyinhibits binding of [3H]a-factor by 50%) were determined from the competition binding curves. The order of potency of unlabeled agonists in competing with [3H]a-factor for binding should parallel the order of efficiency with which these agonists cause their biological effects if all the compounds are interacting with the same receptor. The results demonstratea good correlation of the analogues with respect to K D values and biological activities (Tables I and 111), except for those analogues which bind but are not active. Statistical analysis of the competition curves traced from 10 to 90% competition shows them to be of normal steepness and characteristic of ligand/receptor interactions which are bimolecular, reversible, and obey the law of mass action. Thus, data from the competition studies support the kinetic and saturation binding data discussed earlier that indicate the existence of only one receptor population. This conclusion supersedes our previous suggestion (36) that more than one receptor mediates a-factor responses of agglutination and morphogenesis. Except for [Phe3]a-factor, all tridecapeptides examined in this study had similar binding affinities. Conservative changes at positions 5 and 7 in thetridecapeptide result in no change in binding or biological activity when compared to thenative 1

J

17338

Binding of S. cerevisiae a-Factor Analogues

DHP@DHP” N I P

FIG. 4. Competition for bindingof 100 [3H]a-factor by unlabeled a-factor tridecapeptides. a-Factor and ana90 logue binding t o 4202-15-3 (MATa)cells was performed in competition with [3H] 00 a-factor as described under “Materials andMethods.”Datashownare the 70 means of duplicate determinations, J with each curve representative of two to 60four separate experiments.The symbols t-z representcompetition of [3H]oc-factor 8 5 0 binding by [DHP8,DHP11,Nle’2]trideca- L L peptide (O),native a-factor (A), [Nle”] 40tridecapeptide (W), [Asn6,Arg7]trideca-d 30 peptide (VI, [Phe3]tridecapeptide(4), [octanoyl-Ly~~ltridecapeptide (O), and [acetyl-Lys’ltridecapeptide (0).Bars in20 dicate error. 10 I

IO*

10-8

I

I

I

I

,,,I IO

-’

ANALOGUE

I

10-6

IO-=

I

l l 1 U

I -4

[d

FIG. 5. Competition for bindingof [3H]a-factor by unlabeled des-Trp’dodecapeptides. Binding of the des4202-15-3 Trpl-dodecapeptides to (MATa) cells was performed in competition with [3H]a-factoras described under “Materials and Methods.” Data shown are the means of duplicate determinations, with each curve representative of two t o four separate experiments. The symbols represent competition for [3H]~-factor binding by des-Trp1-[Phe3] dodecapeptide (O),des-Trp’-dodecapeptide (A), and des-Trp’-[Ala3]dodecapeptide (4).Bars indicate error.

ANALOGUE

[M]

pheromone. Replacement of Lys7 by Arg was previously re- interaction of the peptideanalogues with membrane lipid and ported to have no effect on the biological activity of the specific binding to the active site of the receptor (39). It is KO for the tridecapeptides pheromone (38). The replacementof methionine with norleu- possible therefore that the apparent cine in the biologically active [Nle’2]tridecapeptide also re- reflects the increase in the overall hydrophobicity that the comsults in no difference in binding affinity when compared to additional Trp residue contributes to these pheromones the native tridecapeptide. Similar, substitution of dehydro- pared tododecapeptides. Similarly, the 1-kcalloss in stability proline for proline at positions 8 and 11decreased the stability found with [Phe3]a-factor might also be explained on the of the a-factor/receptor interactionby only 180 cal, which is basis of the relative hydrophobicities of these pheromones. Three antagonists were identified in the competition studprobably within the experimental error. At least 1kcal separates theAcO values for most of the des- ies: des-Trp’-[Ala3]dodecapeptide, des-Trp’-[Phe3]dodecapeptide, and des-Trp’-[Cha3,~-Leus]dodecapeptide. The desTrp’-a-factor derivatives from the tridecapeptides analyzed in thisstudy. A changeof slightly over 1kcal in A@ increases Trp’-[Ala3]- and des-Trp’-[Phe3]dodecapeptides showed the corresponding equilibrium constant (KO)by a factor of binding affinities equal to the des-Trp’-dodecapeptide; howexhibited a KD 10. Because the binding affinitiesfor the dodecapeptides and ever, the des-Trp’-[Cha3,~-Leu6]dodecapeptide compound,des-Trp’-[Cha3]domost tridecapeptides tested are separated by a t least a 10-fold 5-fold less thanitsparent decapeptide. None of these analogues had biological activity difference, it appears that the N-terminal tryptophan contributes significantly to the stability of the receptor-ligandcom- when tested in the halo or shmoo assays (Table I). It appears dodecapeptide plex. It should be noted, however, that theKOvalues measured therefore that changes at these positions in in this investigation probablyreflect both the nonspecific analogues strongly influence receptor-mediated biological ef-

Binding of S. cerevisiae a-Factor Analogues

17339

I20

FIG. 6. Competition for binding of [3H]a-factorby unlabeled des-Trp'[Cha3]dodecapeptides. Binding of the des-Trp'-[Cha3]dodecapeptide analogues to 4202-15-3 ( M A T a ) cells was performed in competition with [3H]afactor as described under "Materials and Methods." Data shown are the means of duplicate determinations, with each curve representative of two separate experiments.The s y m b o l s represent competition for [3H]a-factorbinding by des-Trpl-[Cha3]dodecapeptide(A),desTrp1-[Cha3-~-Alag]dodecapeptide (W),

aons-His2

der T , ~ ' CHA'

1

110-

100 -

1

0-AI09

,o-Lau6

L-A1a9

4

90. 80.

= ,J 70 60. 0

;p 4 0

-

.

30-

des-Trp'-[DNS-His2,Cha3]dodecapeptide (V),des-Trp1[Cha3,~-Leu6]dodeca-

20.

peptide (O), and des-Trp1[Cha3,~-Alag] dodecapeptide (4).Bars indicate error.

10

I lo9

I

I

I

,,,,,I

Ida

I

167

I

I

I 1 1 1 1 1 1

I 1 1 1 1 f i I

1

106

IC?

I

I

t

ClSfitl

I

I

I

1d4

U

1d3

[MI

ANALOGUE

TABLEI11 Concentrations for50% inhibition of binding (IC&, binding dissociation constants(KD),and Gibbs standard free energy chanzes (AGO) for a-factor and its analogues Peptide

ICW"

KDb

GO'

M

M

kcal

[Gln5,Lys7]Tridecapeptided [Asn5,Arg7]Tridecapeptide [Nle"]Tridecapeptide

1.4 X 6.5 X 2.3 X

lo-' lo-' lo-'

[DHP8,DHP11,Nle12]Tridecapeptide [Phe3]Tridecapeptide Des-Trp'-dodecapeptide Des-Trp'-[Ala3]dodecapeptide Des-Trpl-[Phe3]dodecapeptide Des-Trp'-[Cha3]dodecapeptide Des-Trp'-[Cha3,o-Leu6]dodecapeptide Des-Trpl-[Cha3,acetyl-Ly~~ldodecapeptide Des-Trpl-[Cha3,0ctanoyl-Lys7]dodecapeptide Des-Trp1-[Cha3,~-Alag]dodecapeptide Des-Trp1-[Cha3,~-Ala9]dodecapeptide

8.5 X 3.1 X 5.7 X 5.5 x 6.5 X 1.5 X 7.5 x 2.8 X 9.1 x

10-~ 10-~ 10-~

Des-Trpl-[DNS-His2,Cha3]dodecapeptide

10-7 10-~

>lo-4 1.8 X 7.1 X

1.3 X 1.5 X 1.4 X 2.3 X 2.1 x 1.9 x 6.8 X 1.3 X 1.2 X 1.5 X 3.4 X 1.7 X 6.5 X 2.1 x

lo-'' lo-'' lo-'' 10-'g

lo-' 10-~ 10-~

loT7

10-~

>10-4 1.9 X 10-7 1.6 X

-10.6 -10.6 -10.6 -10.3 -10.4 -10.4 -9.7 -9.3 -9.3 -9.2 -8.8 -7.8 -7.0 -7.7 -9.1 -7.8

IC,, = the concentration of competitor which inhibits binding of [3H]a-factorby 50%. Values were calculated as the 50% point on a regression line incorporating the points of the appropriate competition curve from Figs. 46. * KD = Ki calculated according to Linden (44). Gibbs standard free energy changes. This is authentic a-factor. e Binding dissociation constants (KO) as determined by Scatchard analysis of the appropriate competition curves from Fig. 4. KDas determined from Scatchard analysis of the [3H]a-factor binding isotherm. See Fig. 3. KDas determined from the measurement of k-1 and k l . See Table 11.

'

fects, but have varied effects on binding affinity. It is interesting that theinfluence of position 3 replacement on biological activity is significantly greater for des-Trp'-dodecapeptide analogs than for the tridecapeptide.At present, we cannot offer a satisfactory explanation for the different structureactivity relationships in the dodecapeptide and tridecapeptide analogs of the a-factor. Previous studies by Baffi et al. (36) have shown that thetamine of lysine is not necessary for biological activity. The competition for bindingshown by des-Trpl-[Cha3,acety1Lys7]- and des-Trpl-[Cha3,0ctanoyl-Ly~~ldodecapeptides (Fig. 4) shows that a free t-amine is not necessary for binding to the receptor. The more hydrophobic [octanoyl-Lys7]analogue binds better than the [acetyl-Lys7]analogue, a trend which may explain its greaterbiological activity.

Oneinactive analogue, des-Trp'-[Cha3,~-Alas]dodecapeptide, did not compete for binding of [3H]a-factor. A similar analogue, des-Trp'-[Cha3,~-Alas]dodecapeptide, competed effectively for binding of [3H]a-factorandexhibited a KO slightly lower than thatfor the des-Trp'-[Cha3]dodecapeptide parent sequence. In previous studies (24), we hypothesized that a type I1 @-turnat the Pro8-Glyg residues of a-factor played an importantrole in itsbiological activity. Recent twodimensional NMR studies havegiven evidence that a-factor can assume such a @-turn both in solution (40) and in the presence of phospholipid vesicle^.^ Numerous theoretical and statistical analyses indicate that type I1 @-turns strongly favor a D-residue or Gly at thei+3 position and aredestabilized by L. A. Jelicks, M. Broido, and F. Naider, unpublished results.

17340

Binding of S. cerevisiae a-Factor Analogues

13. Nakayama, A., Miyajima, T., and Arai, K. (1985) EMBO J. 4 , 2643-2648 14. Dohlman, H.G., Caron, M. G., and Lefkowitz,R. J. (1987) Biochemistry 26,2657-2664 15. Blumer, K. J., Reneke, J. E., and Thorner, J. (1988) J. Biol. Chem. 263,10836-10842 16. Marsh, L., and Herskowitz, I. (1988) Proc. Natl. Acad. Sci. U. S. A. 85,3855-3859 17. Naider, F., and Becker, J. M. (1986) CRC Crit. Reu. Biochem. 2 1 , 225-249 18. Masui, Y.,Tanaka, T., Chino, N., Kita, H., and Sakakibara, S. (1979) Biochem. Bwphys. Res. Commun.86,982-987 19. Hartwell, L. H. (1967) J. Bacteriol. 9 3 , 1662-1670 20. IUPAC-IUB Joint Commission on Biochemical Nomeclature (1985) J. Biol. Chem. 2 6 0 , 14-42 21. Khan, S. A., Becker, J. M., Merkel, G. J., and Naider, F. (1981) Zrzt. J. Pept. Protein Res. 17,219-230 22. Shenbagamurthi, P., Naider, F., Becker, J. M., and Steinfeld, A. S. (1983) J. Chromatogr. 2 5 6 , 117-125 23. Shenbagamurthi, P., Kundu, B., Becker, J. M., and Naider, F. (1985) Znt. J. Pept. Protein Res. 2 5 , 187-196 24. Shenbagamurthi, P., Kundu, B., Raths, S., Becker, J. M., and Naider, F. (1985) Biochemistry 2 4 , 7070-7076 25. Baffi, R. A., Becker, J. M., Lipke, P. N., and Naider, F. (1985) Biochemistry 24,3332-3337 26. Tallon, M.A., Shenbagamurthi, P., Marcus, S., Becker, J. M., Acknowledgments-We wish to thank Charles Murphy and Lisa and Naider, F. (1987) Biochemistry 2 6 , 7767-7774 Siard for the peptide sequencing and amino acid analysis, P. Shen- 27. Tam, J. P., Heath, W. F., and Merrifield, R. B. (1983) J. Am. bagamurthi for synthesizing peptides used in the competition studies, Chem. SOC. 105,6442-6455 Stevan Marcus for technical assistance, and Duane Jenness for 28. Raths, S., Shenbagamurthi, P., Becker, J. M., and Naider, F. providing yeast strains. We also gratefully acknowledge Ellen Dixon (1986) J. Bacteriol. 16%1468-1471 for her assistance with the statistical analysis of the binding data and 29. Moore, S. A. (1983) J. Biol. Chem. 2 5 8 , 13849-13856 John Koontz for enlightening discussions and valuable advice 30. Jenness, D. D., and Spatrick, P. (1986) CeEZ 46,345-353 31. Chvatchko, Y., Howald, I., and Riezman, H. (1986) Cell 46,355throughout this project. 364 32. MacKay, V. L., Welch, S. K., Insley, M. Y., Manney, T. R., Holly, REFERENCES J., Saari. G. C.. and Parker. M. L. (1988) . . Proc. Natl. Acad. Sci. 1. Thorner, J. (1981) in The Molecular Biology of the Yeast SacchaL! S. A. '85,55-59 romyces (Strathern, J., Jones, E., and Broach, J., eds) Vol. I, 33. Boeynaems, J. M., and Dumont, J. E. (1980) Outlines of Receptor pp. 143-180, Cold SpringHarbor Laboratory, Cold Spring Theory, Elsevier Scientific Publishing Co., Inc., New York Harbor, NY 34. De Meyts, P., Bianco, A., and Roth, J. (1976) J.Biol. Chem. 2 6 1 , 2. Trueheart, J., Boeke, J. D., and Fink,G. R. (1987) Mol. Cell. Biol. 1877-1888 7,2316-2328 35. Limbird, L. E. (1986) Cell Surface Receptors: A Short Course on 3. McCaffrey, G., Clay, F. J., Kelsay, K., and Sprague, G. F., Jr. Theory and Methods, Martinus Nijhoff, Boston (1987) Mol. Cell. Biol. 7 , 2680-2690 36. Baffi, R. A., Shenbagamurthi, P., Terrance, K., Becker, J. M., 4. Rose, M. D., Price, B., and Fink, G . R. (1986) Mol. Cell. Biol. 6 , Naider, F. N., and Lipke, P. N. (1984) J. Bacteriol. 158,11523490-3497 1156 5. Duntze, W., Stotzler, D., Bucking-Throm, E., and Kalbitzer, S. 37. Klotz, I. (1982) Science 217,1247-1249 (1973) Eur. J. Biochem. 35,357-365 38. Samokhin, G. P., Lizlova, L. V., Bespalova, J. D., Titov, M. I., 6. Kurjan, J., and Herskowitz, I. (1982) Cell 30,933-943 and Smirnov, V. N. (1979) FEMS Microbiol. Lett. 5,435-441 7. Singh, A., Chen, E. Y., Lugovoy, J. M., Chang, C. N., Hitzeman, 39. Schwyzer, R. (1985) in Peptides Structure and Function (Deber, R. A,, and Seeburg, P. H. (1983) Nucleic Acids Res. 1 1 , 4049C. M., Huvby, V. J., and Koppel, K. D., eds) pp. 3-12, Pierce 4063 Chemical Co., Rockford, IL 8. Ciejek, E., Thorner, J., and Geier, M. (1977) Biochem. Biophys. 40. Jelicks, L. A., Naider, F., Shenbagamurthi, P., Becker, J. M., and Res. Commun. 78,952-958 Broido, M. S. (1988) Biopolymers 27,431-449 9. Masui, Y.,Chino, N., Sakakibura, S., Tanaka, T., Murakami, T., 41. Chou, P. Y.,and Fasman, G. D. (1974) Biochemistry 13,211-245 and Kita, J. (1977) Biochem. Biophys. Res. Commun. 7 8 , 534- 42. Gierasch, L. M.,Deber, C. M., Madison, V., Niu, C., and Blout, 539 E. R. (1981) Biochemistry 20,4730-4738 10. Jenness, D. D., Burkholder, A. C., and Hartwell, L. H. (1983) Cell 43. Venkatachalam, C. M. (1968) Biopolymers 6,1425-1436 35,521-529 44. Linden, J. (1982) J. Cyclic Nucleotide Res. 8,163-172 11. Jenness, D. D., Burkholder, A.C., and Hartwell, L. H. (1986) 45. Brown, N. S., Mole, J. E., Weissinger, A., and Bennett, J. C. Mol. Cell. Biol. 6,318-320 (1978) J. Chrornatogr. 148, 532-535 12. Burkholder, A. C., and Hartwell, L. H. (1985) Nucleic Acids Res. 46. Klanschenz, E., Bienert, M., Egler, H., Pleiss, U., Niedrich, H., and Nikolics, K. (1981) Peptides (N.Y.) 2,445-452 13,8463-8475

L-residues (41-43). The striking loss of binding observed in the des-Trp1-[Cha3,~-Alaa]analogue supports our contention that a type I1 @-bendis an important component of recognition of a-factor by its receptor. The development of a competitive binding assay will allow an analysis of the thermodynamics of agonist and antagonist binding. Most significantly, the results of this paper allow us to begin to dissect the contribution of various side chains and chemical modifications to thebinding of a-factor to itsreceptor. Such information should prove valuable in designing affinity probes for tagging the receptor. It is clear, for example, that tridecapeptides are preferable probes compared to homologous dodecapeptides and that attaching reporter groups to the eamine of Lys7 significantly lowers the binding constant. Labeling of membrane-bound receptors remains a formidable task. Any increase in AG"' bind is undesirable, and an order of magnitude increase in KD can mean the difference between a specifically tagged receptor andamultitude of labeled proteins. Work now in progress is attempting to prepare suitable a-factor derivatives based on the above considerations.

Binding of S. cerevisiae a-Factor Analogues SUPPLEMENTARYMATERIAL

TO:

Peptide analogues compete with the binding of i t r m c e p t o r i n saccharornrcer rerertri+e alpha factor to Susan K . Raths.FledNaider. &&.?.rials

17341

and J e f f m y M. Becker

and M e w

Purification of tritium-labelled .-factor was c a r r i e d Out on I Yaters Chromatograph (Yaterr AIIoCiates, nilford, M) equipped w i t h two Model 510 s o l v e n t d e l i v e r y System. a U6K i n j e c t o r . Model 481 v a r i a b l e wavelengthdetector, Model 680 automated g r a d i e n t c o n t r o l l e r and a Yodel 730 data module. Separations were perfomed on a Waters ulondapakreversed-phase C., column (30 cm x 3.9 m 1.0.. 10 urn p a r t i c l e s i z e ) employing I linear gradient of 25-4oX a c e t a n i t r l l e . 0.025% t r i f l u o r m c e t i c a c i d as themobile phase. Flow r a t e s were t y p i c a l l y 1 . 4 01 1.5 n l / m n I S indicated. The absorbance ofthe column eluant was recordedat 220 "a. When necessary 0.26 m n fractions (360 "1 OF 390"1 depending on the flo* ratel were c o l l e c t e d over the duration of the run (30minuter)using a Phamacia Frac-LOO f r a c t i o n collector(Phamacia Inc.. Piscatatmy, NJ) equlpped withsilicanizedborosilicatetuber. F r a c t i o n s c o n t a i n i n g t r i t i u m - l a b e l l e d species were i d e n t i f l e d by t a k i n g IO u l a l i q u o t s from nt. the fractions and counting them ~n Brays Solution (RerearchPmductsInternational. Prospect. 11) ~n a BeckmanLS7000 s c i n t i l l a t i o n c o u n t e r (Beckman l n r t r m e n t r . Pa10 Alto. CA). . ,A Specrfrc activities were deftnedaccording to the number O f countsassociatedwiththe peak area. Peak areas were compared against two external""labelledNle"tridecapeptide factorstandardinjectlonrforquantitatlon.

TO determine if the ['HI * - f a c t o r Was r a d i o l a b e l l e d at t h e amino Jcid Fasitions was sequenced. I n order to f a c i l i t a t e sequencing J ~ A expected. theHPLC-purifiedpheromne conserve l a b e l . t h e ['HI 0 - f a c t o r was diluted with unlabelled Nlc" o-factor before sequencing. Counting o f a l i q u o t l frm the sequencing f r a c t i o n s i n d i c a t e d n d i m c t i r l t y i n p o r t t i o n 2 (PO* o ft o t a lc o u n t s ) .p s r i t i o n e (43% of countr)po*itisn 11 (27s o ft o t a l ) . 8 and 11 was expected slnce c a t a l y t i c Although the radioactivity associated with rerlduer reductionofdehydmproline should r e s u l t i n ['HI p r o l i n e a t t h e s e p o r i t i o n r . t h e 2 was unexpected and IJS l i k e l y due t o chemical incorporation of tritium into residue t h i s t i d i n ei n a dLhydroprolinr-c~ntlining exchange. S i m i l a r tritim incorporation I d e r i v a t i v e o f GnRH 111 previously reported, and t r l t l a t i o n o f GnRH and m q i o t e n r i n I meleculerresulted i n t r i t i u m i n c o r p o r a t i o n i n t o h l s t l d i n e and subsequent s w c i f i c a c t l v l t i a s of7-9 and 5 Ci/ml,r e r p e c t l v e l y . (16).

"-

The b i o l o g i c a l a c t i v l t y o f t h e p u r i f i e d ['HI "factor was tested u s i q the halo assay forgrowtharrest(Fig.9). AS indiCJted by t h e Clealing o f growthJroundtheCenter dirk, the tritiated a-factor i s W i V a l M t In biologicalactivitytothe Ule"tridecapeptide.PlatingofI .dI tp.plc8di r t MIP',MIP".Wle'~-tridec~peptIde 07 U l e U - ~ - f l C t O rJnd ['HI a - R C t o r On J lm6f C O q n I C alphacells(4202-1-2) I h W no i n h i b i t l o n o f growth (data not s h m ) . RUN N S Y l t l indicate that the

? a i d e SepYeDce Determination Unlabelled tridecapeptide and HPLC p u r i f i e d t r i t i u m - l a b e l l e d . - f a c t o r were subjected to automated sequence analysisbysequential Edman degradation an a Beckman System 89011 Series requenator. The d r i e d HPLC puriftedpeptides were r e c o n s t i t u t e d i n 100 u l 50% aceticacid (PlTC) and appliedtothepolybrene primedglass cup o ft h e requenator.Phenylisoth!ocynste C O U O ~ Iand ~ ~ h e o t a f l u o r a b u t v n c ac>d cleavaoe of the reouential amino-teminal amino acids O E C ; i~nt ~h c~c d p .T h i a l b l i n o n e amino acidderivatives were c o l l e c t e di nt h ef r a c t i o n (PTH) d e r i v a t i v e s by c o l l e c t o r and convertedlntotheircorrespondingphenylthiohydantein treatment w t h aqueous acld.Afteldrying. each sequence cyclefraction was reconstituted i n 50 u l a f 100% methanol. 10 u l o f t h l r were analyzed by HPLC on a HaterrLiquid Chromatography h r n o AcidAnalysis System equipped w i t h a Wisp 7108 i n j e c t o r , two Model 510 pumps. a Model 440 absorbance detector. a Model730 data module and a Waters programable system c o n t r o l l e r . Separatlonr were achlevedusing I 3.9 m x 15 cm ( p a r t i c l e size. 4") Haterr Pico-Tag C., column and gradient conditions I S described previously for PTH amino x t d s ( 4 5 1 . The d e r i v a t i v e s produced i n each c y c l e were quantitatedbycornpariron t o I ret ofexternal standards injectedinnediatelybefore and a f t e r each sequence c y c l e sample was run. An amino acid was placed i n sequence i f the amunt O f thecorrerpondingderivativeincreasedin a Certain sequencing cycle and decreased i n subsequent cycle$.

L'Hl .-Factor

Purification

T l i t i a t e d a - f a c t o r was preparedfromtheparent pheromone OHP'. DHP". Nle" o - f a c t o r as was i n i t i a l l y o u t l i n e d i n t h e methods section.Purityofthe crude,labeledpreparation arrerred by t h i n l a y e r chromatography (data not shown). A t l e a s t n i n e d i f f e r e n t howeve7 the m s t r a d i o l a b e l l e d specnes were resolved by TLC species in the crude preparation. abundant component codgrated with the unlabelled N1e"- tridecapeptide.

Figure 8.

-

reverse phase HPLC. The l i b e l l e d O - f a C t O l was p u r l f l e d from the crude preparation using The maJor Aazopeak (R 16.5 mins, Fig. 7, panel A) corresponds t o a major peak o f n d i o a c t i v i t y retaind (ft'actionp 51-59. Fig. 7 , panel B) by the C, column and c o - e l u t e w i t h Nle"-e-factor.

Growth a r r e s ta c t i v i t yo fp u r l f l e d O i r k 1. Disk2. O i r k 3. D i r k 4. Disk 5 . Disk 6.

['HI 0 - f a c t o r

0.1 ug MI?' OH?" Nle"-Tridecapeptide 0.25 ug OHP' DHP" Nle"-Tlidecapeptide 0.5 ug OHP',bHPLL,i(le"-T~idecappptide I ug DHP'.DHP'',Nle"-Tlidecapept~de 1 ug o - f a c t o r 0 . 1 ug - - f a c t o r The top agar contains I x

Center d i r k . Approximately 0.1 ug ['HI 0 - f a c t o r . 4202-15-3 c e l l s . Themedium i s minimal. incorporation of tritium into the histidine b i o l o g i c a la c t i v i t y .

andtwo

prolines of the

Io'

pheromone d i d not a l t e r

To f u r t h e r assess t h e f e a s i b i l i t y o f u s i n g t h e ['HI m-factor i n binding competition studies, an experiment was done to determine whether the tritium-labelled a-factor interacts w i t h 3 - c e l l s i n a way i d e n t i c a l to i t s unlabelled homolog. Using I x 10' 4202-15-3 cellr/ml. to the arount of labelled aloha-factor V I S varied with the mount of unlabelled a-factor always g i v e t h e r a m totalpeptideconcentration.Binding was deteminedat each o f t h e r a t i o si n d l c a t e di nF i g . 9.The p o i n t s approximate i s t r a i g h t l i n e which i s expected i f b t h an i d e n t i c a l manner. theunlabelled and labelled.-factorsinteractwiththereceptorin

'::

Figure 7. Reverse phase chromatography of ['HI a - f a c t o r . Panel A: P r o f i l e of t h e absorbance (220 nm) f o r HPLC separationof crude ['HI m-factor. Panel B: Radioactivity associated with 0.29 minute fractions taken over the 30 minute HPLC run i n panel A. Panel c: P r o f i l eo ft h e absorbance (220 nm) for HPLC separation ['HI a - f a c t o r . Panel 0: Radioactivity associatedwith 0.29 minute ofpurified fractionstaken over the 30 minute HPLC run i n Panel C. I n panelsA and C, the absorbance i s i n d i c a t e d by t h e s o l i d l i n e and the percent acetonitrile (SA) ofthe r o b l l e phase i f indicatedbythe dashed l i n e .

0

Counts eluted with the void volvlne at t h e p o l a r end o f t h e g r a d i e n t m y be a t t r i b u t a b l e t o ['HI incorporation into the aqueous diluent of the preparation during the cmrcial 16.5 n i n maybe f r a w n t s Of l a b e l l i n g process.Smaller peaks w i t h r e t e n t i o n t i m e s l e r r t h a n the labelled tridecapeptide generatedduringthe process of Catalytic reduction or through m d i o a c t i w decay i n storage. None o ft h e ws m a l l e r peaks were b i o l o g i c a l l y a c t i v e i n t h e haloarray(datanotshorn).

1000

0

Those fractions corresponding to t h e 16.5 minute peak were pooled,concentrated as and r e i n j e c t e d using the sane system f o r HPLC reparation. described I n the wthods section The p r o f i l e s f o r e l u t i o n s f r a d i o a c t i v i t y and220 nm absorbingmaterialare s h m i n panels C and 0. respectively.ofFig. 7. The peak o f r a d i o a c t i v i t y i n panel D corresponds t o t h e 14.88minute peak o f absorbance i n panel C. These resultsalongwiththefactthatthe unlabtlled parent compound. Nle" trldecapcptide co-eluted with both peaks indicates that ['HI .-factor was radiochemically pure. The s p e c i f i c x t i v l t y o f ['HI .-factor was calculated based on the total counts associatedwlththecorresponding A,, peak area. T r o 10 ug standards of unlabelledNle" peak ares c m p w i r o n s and subsequent tridecapeptide were injected separately for q u m t i t a t i o n . The s p e c i f i c a c t i v i t i e s v a r i e d s l i g h t l y between batchpurificationsbutIlwws equaled between 8 2nd IO Cl/ml.

0 Ratio vnloballad Letellad

Figure9.

400 4 I

800 1200 1600 2 x) TOTAL C P M 1 2 0 p L

3 2

2 3

1 4

0 S

['HI .-factor interactswiththeo-factorreceptor In a way i d e n t i c a lt oi t s u n l a b e l l e dh m l o g . The a m u n to f ['HI e - f a c t o r bound ( c p l ) i s graphed vs. t h e t o t a l ['HI *-factor present. The r a t i o o f v n l a b e l l e d a - f a c t o r to ['HI afactor for each p o i n t i s i n d i c a t e d below t h e graph.