INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

There has been little work on the binding characteristics of microbial cell adhesion proteins, which can function under a variety of conditions not encountered in cell adhesion systems of multicellular organisms (28). Sexual cell adhesion in the yeast Saccharomyces cerevisiae is mediated through two interacting cell wall glycoproteins, a-agglutinin and -agglutinin, which are expressed on haploid a and  cells, respectively (29, 41). These proteins are expressed at low levels constitutively, and gene expression and cell surface concentration are upregulated in response to a sex pheromone produced by the opposite mating type. The pheromones do not change the binding affinity (40).

a-Agglutinin consists of two subunits. The binding subunit Aga2p is a glycopeptide containing 69 amino acid residues and about 20 O-linked oligomannosyl chains (3). Aga2p carries the adhesive domain, and its C-terminal 10 amino acids (sequence GSPINTQYVF ) can act as a ligand for -agglutinin (2). A pair of disulfide bonds links Aga2p to the anchorage subunit of a-agglutinin Aga1p (2, 35). Aga1p itself is a highly O-glycosylated protein of 725 amino acid residues including an N-terminal secretion signal and a C-terminal signal for addition of a glycosyl phosphatidylinositol anchor (GPI). Aga1p anchors Aga2p onto the cell surface but has no direct role in binding (35, 35a).

-Agglutinin, the other cell adhesion molecule responsible for agglutination, is the product of the AG1 (SAG1) gene. The N-terminal half of -agglutinin is homologous to the immunoglobulin superfamily and contains the binding site for a-agglutinin (4, 16, 42). This region is -sheet rich and shows a high degree of conformational flexibility (45, 46). Residues near His²⁹² and Lys¹⁵⁴ have been implicated in binding (2, 4, 6, 45). The C-terminal half of -agglutinin anchors the protein and consists of a highly glycosylated "stalk" of about 300 amino acids. This is followed by a modified GPI anchor that covalently attaches the molecule to 1,6-glucan, a cell wall polysaccharide (31, 32, 42). These features of the -agglutinin sequence are similar to those of several Candida albicans cell adhesion proteins that mediate host-pathogen interactions (11-13, 17, 18).

The agglutinins bind tightly to each other at low to moderate ionic strengths and with a pH optimum of 5.5. Binding of ¹²⁵I--agglutinin fragments to cell-bound a-agglutinin showed a specific and complex interaction consistent with at least two states, weak and tight. Weakly bound ¹²⁵I--agglutinin was removed from the a cells by washing. The tight state had a K_D of about 10⁹ M and was essentially irreversible, implying a low rate of dissociation (30). Formation of this state was cold sensitive.

We now describe a study of the binding interaction based on highly purified recombinant agglutinins. A surface plasmon resonance (SPR) analysis confirmed the original estimates of K_D and together with circular dichroism (CD) showed a conformational basis for complex binding kinetics. The binding kinetics demonstrated a basis for the cold sensitivity of sexual agglutination.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Materials. The following reagents were obtained from Sigma Chemical Company, St. Louis, Mo: cycloheximide, p-chloromercuribenzoic acid (PCMB), -methylmannoside, and bovine serum albumin (BSA). Protein standards, Bio-Gel P-60, and Bio-Gel P-100 were purchased from Bio-Rad (Richmond, Calif.). Restriction enzymes were from New England Biolabs. Endoglycosidase H was from Boehringer Mannheim. Plasmon resonance experiments were performed on BIACORE X, which together with C₁ sensor chips, N-hydroxysuccinimide (NHS), N-ethyl-N'-(3-dimethylaminopropyl)-carbodiimide hydrochloride, and 1 M ethanolamine (pH 8.5) were obtained from Pharmacia Biosensor AB (Uppsala, Sweden). Polyacrylamide gel electrophoresis (PAGE) reagents and standards were from Bio-Rad.

Purification of Aga2p. W303 diploid cells containing plasmid YEpPGK-AGA2 constitutively expresses Aga2p under the control of the PGK promoter. Cells were grown in synthetic medium without uracil to 4 × 10⁷/ml and harvested. Cell culture supernatants were then collected and concentrated 10-fold through a Millipore filter with a 10-kDa molecular-mass cutoff. Concentrated supernatant was further dialyzed against 20 mM sodium acetate (pH 5.5)-1 mM EDTA, lyophilized, and resuspended in the same buffer with 10% (vol/vol) glycerol. Aga2p was purified using a Bio-Gel P-60 size exclusion column equilibrated in 20 mM sodium acetate, pH 5.5. Because of its extensive glycosylation, Aga2p has an apparent molecular mass of 33 kDa on sodium dodecyl sulfate (SDS)-polyacrylamide gels stained with Coomassie blue (Fig. 1).

View larger version (65K): [in this window] [in a new window]   FIG. 1.   SDS-PAGE analysis of -agglutinin and Aga2p. Protein samples were electrophoresed in SDS on a 10% polyacrylamide minigel and stained with Coomassie blue. First lane, molecular weight (MW) standards (phosphorylase b, 97,400; BSA, 66,200; ovalbumin, 45,000; carbonic anhydrase, 31,000; soybean trypsin inhibitor, 21,500; lysozyme 14,400); second lane, Aga2p, glycosylated protein with apparent molecular mass of 33 kDa; third lane, -agglutinin, glycosylated protein with an apparent molecular mass of 45 kDa; fourth lane, BSA, 1 µg.

Purification of -agglutinin. -Agglutinin containing residues 20 to 351 (-agglutinin^20-351) was purified from the cell culture supernatant (4 liters) of yeast strain L21 (MAT  ade2-1 his3-11,15 leu2-3,112 trpl-1 ura3-1 can1-100 ag1-3) containing pPGK-AGl³⁵¹. A DEAE-Sephadex A-25 column and a Bio-Gel P-60 size exclusion column were used for the purification (4). Partially deglycosylated -agglutinin^20-351 had an apparent molecular mass of 45 kDa on Coomassie blue-stained SDS-polyacrylamide gel and was of 99% purity by scanning densitometry (Molecular Dynamics) (Fig. 1).

PAGE. The protein samples were electrophoresed in SDS- or non-SDS-polyacrylamide gels with 10% running gels and 4% stacking gels (26).

Anti--agglutinin antibody preparation. Anti--agglutinin antibody was produced by injecting purified protein (100 µg, twice) in Hunter's Titermax into rabbits. The sera were then collected 10 to 25 weeks after the initial injection and were absorbed twice against washed, heat-killed a cells.

Separation of -agglutinin aggregates and immunodetection. A sample of purified -agglutinin was thawed and filtered through a Bio-Gel P-100 column in 100 mM sodium acetate, pH 5.5. The samples collected from the column were electrophoresed in non-SDS- and SDS-10% polyacrylamide gels and then transferred to a nitrocellulose membrane by electrophoresis. The blot was first incubated in 10 mM Tris-140 mM NaCl (pH 7.4; buffer B), with 3% gelatin for 30 min and then in anti--agglutinin antibody (diluted 1:500) for 1 h. Afterwards, the blot was washed and incubated in anti-rabbit immunoglobulin G peroxidase conjugate (diluted 1:1,000) in buffer B with 1% gelatin for 1 h. The blot was then washed and visualized with 4-chloro-1-naphthol in 10 ml of cold methanol and 3% hydrogen peroxide in 50 ml of buffer B. The fractions containing no aggregates were used for BIACORE experiments.

SPR. The detection of binding or dissociation in SPR is based on the change in the refractive index due to association of a soluble analyte to an immobilized ligand. The resulting signal is proportional to the amount of protein bound, and 1,000 resonance units correspond to 1 ng/mm² (20). The flat carboxymethylated surface of a C₁ sensor chip was first cleaned with 10 µl of 0.1 M glycine-NaOH and then activated by 35 µl of a solution containing 0.2 M N-ethyl-N'-3-aminopropyl carbodiimide and 0.05 M NHS at a flow rate of 5 µl/min. Then, two 25-µl aliquots of purified Aga2p (5 µg/ml) in 100 mM phosphate buffer were injected over the surface at a flow rate of 5 µl/min. The remaining unreacted NHS-ester groups were deactivated by injection of 25 µl of 1 M ethanolamine-HCl, pH 8.5, and the surface was further blocked by injecting if twice with 25 µl of 100-µg/ml BSA in HBS buffer (10 mM HEPES and 150 mM NaCl, pH 7.4). The flow rate was 5 µl/min.

Isotopic labeling of -agglutinin with ¹²⁵I-Bolton Hunter reagent. Purified -agglutinin (0.1 mg/ml) was dialyzed against 200 mM sodium phosphate buffer, pH 8.5, overmight. The protein (20 µg) was reacted with 250 µCi of ¹²⁵I-Bolton Hunter reagent (New England Nuclear Corp.) for 3 h at 0°C. The reaction mixture was diluted 20-fold in the reaction buffer (100 mM sodium acetate [pH 5.5], 1 µM CaCl₂, and 1 µM MnSO₄). The diluted sample was applied to a 1-ml prepacked lentil lectin-agarose column, which was preequilibrated with reaction buffer. The protein was eluted with 1 M -methylmannoside and 500 mM sodium chloride in 100 mM sodium acetate, pH 5.5. The labeled -agglutinin was stored at 4°C in the eluting buffer supplemented with 10% glycerol and 1 mg of BSA per ml.

Binding of ¹²⁵I--agglutinin to yeast cells. Various concentrations of ¹²⁵I--agglutinin were incubated with 2 × 10⁶ X2180-1A (MATaSUC2 mal mel gal2 CUP1) or X2180-1B (MAT SUC2 mal mel gal2 CUP1) cells in a total volume of 250 µl on a rotary shaker at 200 rpm at 25 or 0°C for 90 min, the time required to reach equilibrium. The buffer used for this study was 100 mM sodium acetate, pH 5.5, supplemented with 1 mg of BSA per ml, 1µg of cycloheximide per ml, 1 µM p-chloromercuribenzoic acid (39). After the incubation, cells with bound -agglutinin were centrifuged, the supernatants were aspirated, and the cells were then washed three times. The bound ¹²⁵I--agglutinin was counted in a Compugamma 1282 (LKB Wallac Inc.), and specific binding was determined as the difference in the levels of binding between MATa and MAT cells under identical conditions. The nonspecific binding to MAT cells was not saturable and was similar to noncompetable binding to MATa cells, as previously reported. (39). Equilibrium constants were determined from Scatchard analysis of the binding.

Concentration determination of ¹²⁵I--agglutinin. Samples of ¹²⁵I--agglutinin (10 µl) were incubated with 2 × 10⁶ a or  cells in a 200-µl total volume to obtain the maximum specific binding capacity of a cells. Specific binding was then determined in a competition assay between ¹²⁵I--agglutinin (5 µl) and different concentrations of unlabeled -agglutinin of known concentration. After incubation, the cells with bound ¹²⁵I--agglutinin were washed and counted. The concentration of ¹²⁵I--agglutinin was calculated to be 0.105 µM in the stock solution, based on the concentration of unlabeled -agglutinin giving 50% inhibition of binding (1).

Peptide synthesis. A peptide containing the C-terminal 10 residues of Aga2p was synthesized by the solid-phase method using fluorenylmethoxycarbonyl chemistry on an Applied Biosystems automated model 432A peptide. The peptide resins were treated with 80% trifluoroacetic acid (TFA)-5% water-5% ethanedithiol-10% thioanisole at room temperature for 2 h to cleave and deprotect the peptides, which were then precipitated and washed in cold methyl t-butyl ether and collected by centrifugation at 25°C.

Agglutination assay. S. cerevisiae wild-type haploid strains X2180-1A (MATa SUC2 mal mel gal2 CUP1) and X2180-1B (MAT  SUC2 mal mel gal2 CUP1) were used for bioassays. a cells and  cells were grown separately in minimal medium to 2 × 10⁷ cells per ml, and a cells were treated with the sex pheromone -factor as described previously (41). These cells were harvested and washed in 100 mM sodium acetate, pH 5.5, at 25°C. -Agglutinin was incubated with a cells on a rotary shaker at 25°C for 90 min, and  cells were then added. The activity of -agglutinin was determined by its ability to inhibit the agglutination of a cells (41), with 1 U being the amount of protein needed to inhibit agglutination by 10%.

CD spectroscopy. Far-UV CD spectra were recorded on a Jasco J-710 spectropolarimeter equipped with a thermo-regulated cell holder with a 0.05-cm path length (HELLMA). Each spectrum represents the average of 10 spectra taken at 0.5-nm intervals from 250 to 200 nm. The spectra were corrected by subtraction of the appropriate buffer baseline spectra and smoothed by Jasco Series 700 software. Each experiment was repeated three times. CD data presented here were analyzed by the self-consistent method (SELCON) (36, 37).

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Binding of ¹²⁵I--agglutinin to intact cells. As indicated in Fig. 2, the binding of ¹²⁵I--agglutinin to a cells at 25 and 0°C showed only slight differences in affinity or site numbers. There was little nonspecific binding to cells of mating type . When the protein concentration was 5 nM or higher, the binding approached saturation. The inset of Fig. 2 shows Scatchard plots of the binding, with cells displaying a K_D of 6.34 × 10¹⁰ M at 0°C and 8.6 × 10¹⁰ M at 25°C, values similar to that previously reported (30). There were about 7.5 × 10⁴ -agglutinin molecules bound per a cell at saturation. For samples incubated at either temperature, washing the cells before counting did not reduce the specific binding significantly. Therefore, no weak interactions were observed at either temperature. We speculate that the weak and cold-sensitive binding seen in a previous study might have been due to low-affinity binding of partially active fragments of -agglutinin (30).

View larger version (24K): [in this window] [in a new window]   FIG. 2.   Binding of purified 125I--agglutinin to a cells. Increasing concentrations of 125I--agglutinin were incubated with 2 × 106 MATa cells or MAT cells for 90 min. After a thorough washing, levels of bound 125I--agglutinin were determined. Binding isotherms are shown for 25°C (circles) and 0°C (triangles). Binding to MATa cells is shown with filled symbols; binding to MAT cells is shown with open symbols. Error bars represent the ranges for results of triplicate samples. (Inset) Scatchard plots for specific binding at each temperature.

SPR analysis of binding kinetics. In equilibrium binding experiments with radioactively labeled -agglutinin, low temperature had little effect on the binding isotherm. However, previous studies showed that low temperature does reduce cellular agglutinability both in vivo and in vitro (41, 43). SPR was used to obtain the k_off and the k_on and to detect the effect of low temperature on the interaction of the agglutinins in real time.

5

1

1

1

9

5

1

1

1

1

1

4

10

View larger version (17K): [in this window] [in a new window]   FIG. 3.   SPR analysis of binding of -agglutinin to Aga2p. Aga2p was coupled to a C1 pioneer chip through amine chemistry. Aliquots (100 µl) at various concentrations of -agglutinin were then injected over the sensor chip at 20 µl/min in 10 mM MES buffer plus 150 mM NaCl, pH 5.9. (A) Binding at 20°C; (B) binding at 10°C.

Effects of protein aggregates. Protein aggregates may sometimes bind with high affinity, and dissociation rates decrease in geometric proportion to the number of the binding sites in the aggregates (24, 44). To correct the estimate of k_off for the presence of aggregates, gel-filtered -agglutinin was prepared and injected into the BIACORE instrument. This material had a k_on of (5.6 ± 2.2) × 10⁴ M¹, similar to that of the experiment shown in Fig. 3, and a higher k_off of (2.6 ± 0.6) × 10⁴ s¹, corresponding to a K_D of (5.5 ± 2.6) × 10⁹ M.

Conformational changes of the agglutinins induced by a-agglutinin peptide binding. Sigmoidal association kinetics and low k_on and k_off are characteristics of binding dependent on conformational change in the ligands (induced-fit models [25]). A CD analysis of binding of -agglutinin binding to an a-agglutinin peptide tested for conformational changes on binding.

6

5

View larger version (19K): [in this window] [in a new window]   FIG. 4.   Far-UV CD of -agglutinin induced by binding of the a-agglutinin peptide. Far-UV CD spectra of -agglutinin were measured at 25°C in 10 mM sodium acetate buffer, pH 5.5 (---). Also shown is a difference spectrum of -agglutinin incubated with the Aga2p C-terminal peptide, with the CD contribution of the peptide subtracted (····). deg., degrees.

Effect of reduced temperature on the conformation of -agglutinin. To determine if cold temperature affected the secondary structure of -agglutinin, far-UV CD spectra of -agglutinin were measured and compared at 25 and 0°C. As shown in Fig. 5, CD spectra at both temperatures showed a major band at 217 nm, a contribution of peptide chromophores in -conformation. However, at 0°C, the magnitude of the negative band at 217 nm was intensified and broadened at lower wavelengths, indicating a possible increase of -sheet structure. In addition, at 0°C, the spectrum was less negative between 227 and 237 nm. The protein appeared to be more structured at low temperature than at room temperature: the total -sheet content increased from 47.8 to 54.4%, and the random structure dropped from 36.5 to 26.1%. Such a change corresponds to an addition of about 24 to 34 residues to regions with periodic secondary structures (-helix or -sheet).

View larger version (18K): [in this window] [in a new window]   FIG. 5.   Effect of cold temperature on far-UV CD of -agglutinin. Far-UV CD spectra of -agglutinin were measured in 10 mM sodium acetate buffer, pH 5.5, at 25°C (---) and after incubation at 0°C for 20 min (····). deg., degrees.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Complexity of binding. Kinetics and secondary-structure analysis of interacting yeast sexual agglutinins showed complexity and the characteristics of an induced-fit interaction (25). The binding parameters obtained in various experiments yield K_Ds of 1 × 10⁹ to 5 × 10⁹ M, corresponding to a binding energy of 11 to 12 kcal/mol. The uncertainty in K_D results from the uncertainty in the k_off, which is close to the limit of reliability of SPR, and indeed equilibrium binding gives a K_D corresponding to the tighter binding value.

Effect of low temperature on protein flexibility. Incubation at 0 to 10°C results in poor agglutination of intact cells (41, 43). In SPR experiments (Fig. 3), there was also a marked decrease in the initial association rate of the agglutinins at low temperatures. This result can be interpreted as reduced flexibility in one or both of the components. There is direct evidence for such reduced flexibility in -agglutinin at reduced temperatures. This protein showed a 7 to 10% increase in the fraction of the periodic secondary structure (-helix and -sheet) and a concomitant decrease in the fraction of the aperiodic secondary structure. The increase in the fraction of the periodic secondary structure encompasses 24 to 34 residues, far greater than the standard deviation for secondary-structure assignments (about 1.3% or 4 residues). This change represents a decrease in the inherent flexibility of -agglutinin (45, 46), because the conformations of residues in sheets and helices are stabilized by backbone hydrogen bonding, as well as by side chain interactions. However, the binding characteristics implied that, at low temperatures, the rate-limiting step was a conformational change in the ligand a-agglutinin.

Conformational changes on binding. The conformational change observed on binding includes removal of about 24 amino acids from less structured aperiodic conformations to helix or sheet. There are 10 residues in the a-agglutinin peptide used as the ligand, so there is an approximate minimum of 14 residues of -agglutinin whose secondary-structural state is altered by binding. There is a precedent for this type of interaction: PDZ domains bind the C-terminal regions of their ligands, with the ligand forming an extra -strand in a sheet. Like the agglutinin system, the ligand-receptor complex is more structured than the sum of the individual components. Also, like in the agglutinin system, binding is tight and dissociation is extremely slow (5, 7).

A model of the interaction of the agglutinins at different temperatures. A model of the interaction between the agglutinins based on binding data and secondary-structure analysis is shown below: inhibited at low temperature +[] [a'] [a]  [a·]

Similar binding systems. The above model of binding-induced conformational changes of the agglutinins is similar to the induced-fit model of interaction of substrate and enzyme, where the active site of the enzyme assumes a shape complementary to that of the substrate after it is bound (25). Well-characterized examples include interactions of antigens with antibodies. For example, a structural rearrangement next to the helical binding site of a tobacco mosaic virus protein occurs during the high-affinity interaction between the protein and its antibody (9, 34). Slow conformational changes of human immunodeficiency virus core protein p24 are involved in the high-affinity interaction of the protein to its own monoclonal antibody (14). Conformational changes of the integrin _IIb3 (platelet GPIIb-IIIa) triggered by binding of Arg-Gly-Asp sequences in fibrinogen generate a high-affinity binding state of integrin (8). Using SPR, Huber also studied the interaction of _IIb3 with fibrinogen and found an initial low-affinity interaction leading to a higher-affinity state (19). Conformational changes of the cell adhesion protein thrombrospondin are induced by binding to the receptor CD36 (GPIIIb or GPIV) (27).

Biological consequences of binding. In the yeast sexual agglutination system, the situation in vivo is that cellular aggregation is irreversible. Cell surface concentrations of the agglutinins are about 10³ M after their expression has been upregulated by sex pheromones (35a). At this concentration, strong adhesive bonds between agglutinating cells would form within a few seconds, as has been observed previously (38). A dissociation rate of 10⁴ s¹ would mean that two singly tethered cells likely dissociate in several hours, about the time two cells take to fuse. The k_off values for bivalent systems are the products of the univalent rates. So, if two cells were attached through two adhesive bonds, the dissociation of most pairs would take 10⁸ s (more than a year). Thus, in contrast to the transient interactions that characterize cell interactions in the immune system, yeast sexual agglutination makes adhesive bonds that are irreversible within biologically relevant times.