Cell, Vol. 63, 1039-1051,
November
30, 1990, Copyright
0 1990 by Cell Press
Courtship in S. cerwisiae: Both Cell Types Choose Mating Partners by Responding to the Strongest Pheromone Signal Catherine L. Jackson and Leland Department of Genetics, SK-50 University of Washington Seattle, Washington 98195
H. Hartwell
Summary We demonstrate that during the courtship stage of conjugation, S. cerevlsiae a cella choose the a cell producing the highest level of pheromone from among potential matlng partners. From this result and that for a cells we conclude that both a and a cells act as signaling cells during courtship, that both cell types respond by dlscrlminatlng different levels of signal, and that the signals are the mating pheromones. Responding cells that are supersensitive to slgnal fall to discrimlnate pheromone-producing from nonproduclng cells to an extent that depends on their degree of supersensltlvlty. We propose that partner selection in S. cerevlsiae results from polarized morphogenesis of a responding cell in the direction of hlghest pheromone concentration and that cells defective in dlscriminating this gradient execute a default pathway in which an adjacent cell is selected at random. Introduction Conjugation in Saccharomyces cerevisiae involves a series of cell-cell interactions (for reviews see Nasmyth and Shore, 1987; Cross et al., 1988; Herskowitz, 1989). When haploid a and a cells first come into contact, they detect each other’s presence by responding to the mating hormone (a- or a-pheromone, respectively) produced by cells of the opposite mating type. The responses include induction of expression of gene products that facilitate mating (such as the cell type-specific agglutinins and the FUS7 protein necessary for cell fusion) and arrest of the cells in Gl (Cross et al., 1988). The agglutinins mediate specific adhesion of a and a cells, which results in a stable association between cells of opposite mating type (Watzele et al., 1988; Lipke et al., 1989). Once the cells have come into close contact, each cell becomes committed to a particular mating partner-a step we have termed courtship (Jackson and Hartwell, 1990). After mating partners have been selected, the final events of cell and nuclear fusion take place at the site where a pair of cells are apposed. The cell walls fuse, then are broken down, forming a pore through which the cell membranes and then the nuclei fuse (Byers and Goetsch, 1975; Byers, 1981). We are interested in the question of how a cell chooses a mating partner during courtship. Using the competition mating assay, in which wild-type a cells had a choice of two a cell mating partners, we demonstrated previously that an a cell can choose between potential a cell mating partners (Jackson and Hartwell, 1990). In this study, we use the competition mating assay to show that the a cell
can also make a choice of mating partner from among adjacent a cells. How do cells communicate to choose one from among several possible mating partners? A strong circumstantial case can be made for a central role of mating pheromones in this process. Although pheromoneless a cells (Kurjan, 1985) and pheromoneless a cells (Michaelis and Herskowitz, 1988) could be stimulated to undergo a mating response, these cells were not chosen as mating partners by cells of the opposite mating type, even in the presence of pheromone-producing wild-type a and a cells, respectively. Hence, even though both a- and a-pheromones are diffusible, pheromone produced by one cell is insufficient to allow mating of an adjacent nonproducing cell, and mating is restricted to those cells that produce pheromone. The authors suggested that a cell must present its own pheromone in order to be chosen as a mating partner by a cell of opposite mating type. This requirement for presentation may indicate that pheromone is the signal for partner selection, or that it is necessary for some later step in mating. The results of Bender and Sprague (1989) suggest the former is true for a cells, since they showed that a haploid cell need only express a-pheromone receptor and a-pheromone in order to mate with an a cell. However, the mating efficiency observed in these experiments was at least 5-fold lower than that of wild-type a and a cells, leaving open the possibility that some other signal is necessary for efficient choice of mating partner. We have examined the role of pheromones in partner selection. We showed previously that an a cell chooses for a mating partner the adjacent cell producing the highest level of a-pheromone (Jackson and Hartwell, 1990). In the present study, we show that an a cell chooses the adjacent a cell producing the highest level of a-pheromone. These results indicate that the mating hormones, a- and a-pheromone, are the signals for partner selection and that both cell types act as responders (by responding to the relative concentrations of complementary pheromone) and signalers (by secreting the pheromone signal) during courtship. We suggest that courtship results in the polarized morphogenesis of the responding cell in response to a gradient of pheromone in its environment. Because of the powerful genetic, biochemical, and cell biological techniques that can be applied to S. cerevisiae, studies of courtship in S. cerevisiae should provide insights into the mechanisms by which other eukaryotic cells generate cellular asymmetry in response to environmental signals. Pertinent examples are the polarized growth of nerve cells stimulated by nerve growth factor and other molecules (Dodd and Jessell, 1988), migrations of cells and axons in Caenorhabditis elegans (Hedgecock et al., 1990), migration of T-lymphocytes and their interaction with target cells during an immune response (Young and Cohn, 1988; Springer, 1990), and chemotaxis in Dictyostelium discoidium and polymorphonuclear leukocytes (Devreotes and Zigmond, 1988; Devreotes, 1989).
Cell 1040
COMPETITION
A
MATING ASSAY
Non-competition
Competition
Good competition
Index:
B
DISCRIMINATION
Discrimination
Randomness
0
ASSAY Random Choice
Index:
1
Figure 1. Schematic Diagrams of the Competition Mating and Discrimination Assays Used to Assess the Ability of Cells to Choose Mating Partners during Courtship In both assays, the ratio of responder x cells (represented by open circles) to signaling y cells (solid or stippled circles) is low so that each x cell is surrounded by y cells. The arrows represent potential interactions between the responder x and signaling y cells during courtship. (A) In the competition mating assay, only mating between an x and a target y cell (yr, solid circles) is measured; the ability of the challenger y cells (yc, stippled circles) to compete-that is, to inhibit mating between an x and a y, cell by attracting the x cell-is monitored. The ability to compete is quantified by computing the Cl (see Experimental Procedures). A noncompetitive ys cell fails to attract the responder x cell (resulting in a low Cl), whereas a yc cell that is a good competitor attracts an x cell as efficiently as a yc cell does (resulting in a Cl of 1). If a good cumpetitor yc strain is mating proficient, a zygote is formed between the responder x and yc cells; if not, the x and yc cells undergo the courtship interaction, but then fail to complete the process of diploid formation. (B) In the discrimination assay, the relative mating efficiency of the responder x cells with the two different types of signaling y cells is monitored. The signaling y cells produce different levels of pheromone, the courtship signal. Cells producing the higher level of pheromone are solid circles, and cells producing the lower level (usually no pheromone) are stippled. The two most extreme cases are shown: a responder x cell that discriminates perfectly chooses exclusively a cell producing the higher level of pheromone as a mating partner (re(,:esented by a randomness index of 0, see text), whereas a cell completely defective in discriminating the two types of y cells chooses a mating partner at random (represented by a randomness index of 1).
assay and the discrimination assay (Figure 1). in both assays, one ceil type (MATx or x) is given the choice of two different mating partners of opposite mating type (MA? or y). Since each assay monitors the ability of an x cell to choose one of the two types of y ceil as a mating partner, the x ceil is referred to as the responder and the y ceil the signaler in mating partner selection. In both assays, the ratio of x to y cells is low so that each x ceil is surrounded by y ceils. Hence, the responding role of the x cell can be assessed independently of the response of the y cells, since ail of the y cells surrounding a given x cell will respond only to that x ceil and therefore have no choice to make. In the competition mating assay, the two types of signaling y ceils are termed target and challenger. Diploids produced between x and only one of the potential y partners (the target ceil) are monitored. The role of the challenger ceil in producing the signal for mating partner selection is assessed by the effect it has on mating of the x with the target y cells. in this assay the x cell and the target y cell are normally wild type and the genotype of the challenger is varied to assess the role of various genes in signal production. The advantage of this assay is that the challenger does not need to be mating proficient to have a detectable effect. That some challenger cells do compete with a wild-type target y cell for mating with a responding x ceil while others do not establishes that a responder ceil can make a choice between potential partners. The finding that some mating-defective mutants can compete demonstrates that partner selection is distinct from and a prerequisite to conjugation (Jackson and Hartwell, 1990; see below). In the discrimination assay, the relative efficiency of mating of x ceils with the two potential y partners is measured. This assay is used to monitor the ability of the responding x ceil to discriminate between potential y partners. One y partner normally produces the signal for mating partner selection and the other does not (or produces a reduced amount). The advantage of this assay is that small defects in the ability of the responding cell to discriminate mating partners can be measured sensitively, since mating with one or the other y strain can be measured over a 104-fold range. To draw conclusions about the ability of the responding ceil to choose partners, it is necessary to show that the two y cells have the same intrinsic capacity to mate, independent of their ability to produce signal. This result is established by showing that there are conditions under which the poorer signaler is able to mate as efficiently as a wild-type strain. In several experiments, the concentration of a-pheromone by a cells is varied by using strains defective in one or both of the a-pheromone structural genes, MFa7 and MFa2. An mfal strain produces approximately 20-fold less a-pheromone not shown),
than
a wild-type mfa2
a strain
(Kurjan,
strain produces
1985;
data
Results
and an mfal omone (Kurjan, 1985).
no a-pher-
Experimentsi Design We use two assays to reach conclusions about interceiiuiar communication during mating: the competition mating
a Ceils Can Choose Among Potential a Mating Partners We showed previously that the a cell plays the role of re-
Mating 1041
Table
Partner
Selection
1. Competition
in S. cerevisiae
Mating
Assay
with Wild-Type
a
Responding
Theoretical a, Strait?
at Strains
rc
MFal MFa2 mfal mfa2 ste3 sfe3
MFa7 MFa2 MFal MFa2 MFal MFa2 mfal MFa2
0.22 0.24 0.24 0.25
f f + f
0.037 0.015 0.039 0.016
(4) (5) (3) (3)
Cells Cld for CI~ a
Noncompetitor
Good
0.24 0.25 0.25 0.27
1 1 1 1
f 0.032 f 0.016 f 0.034 f 0.014
Competitor
Observed 1.17 0.29 0.73 2.26
2 f f f
Mating
Cle 0.33 0.03 0.34 0.90
(4) (5) (3) (3)
Efficiency
24.2 f
November
30, 1990, Copyright
0 1990 by Cell Press
Courtship in S. cerwisiae: Both Cell Types Choose Mating Partners by Responding to the Strongest Pheromone Signal Catherine L. Jackson and Leland Department of Genetics, SK-50 University of Washington Seattle, Washington 98195
H. Hartwell
Summary We demonstrate that during the courtship stage of conjugation, S. cerevlsiae a cella choose the a cell producing the highest level of pheromone from among potential matlng partners. From this result and that for a cells we conclude that both a and a cells act as signaling cells during courtship, that both cell types respond by dlscrlminatlng different levels of signal, and that the signals are the mating pheromones. Responding cells that are supersensitive to slgnal fall to discrimlnate pheromone-producing from nonproduclng cells to an extent that depends on their degree of supersensltlvlty. We propose that partner selection in S. cerevlsiae results from polarized morphogenesis of a responding cell in the direction of hlghest pheromone concentration and that cells defective in dlscriminating this gradient execute a default pathway in which an adjacent cell is selected at random. Introduction Conjugation in Saccharomyces cerevisiae involves a series of cell-cell interactions (for reviews see Nasmyth and Shore, 1987; Cross et al., 1988; Herskowitz, 1989). When haploid a and a cells first come into contact, they detect each other’s presence by responding to the mating hormone (a- or a-pheromone, respectively) produced by cells of the opposite mating type. The responses include induction of expression of gene products that facilitate mating (such as the cell type-specific agglutinins and the FUS7 protein necessary for cell fusion) and arrest of the cells in Gl (Cross et al., 1988). The agglutinins mediate specific adhesion of a and a cells, which results in a stable association between cells of opposite mating type (Watzele et al., 1988; Lipke et al., 1989). Once the cells have come into close contact, each cell becomes committed to a particular mating partner-a step we have termed courtship (Jackson and Hartwell, 1990). After mating partners have been selected, the final events of cell and nuclear fusion take place at the site where a pair of cells are apposed. The cell walls fuse, then are broken down, forming a pore through which the cell membranes and then the nuclei fuse (Byers and Goetsch, 1975; Byers, 1981). We are interested in the question of how a cell chooses a mating partner during courtship. Using the competition mating assay, in which wild-type a cells had a choice of two a cell mating partners, we demonstrated previously that an a cell can choose between potential a cell mating partners (Jackson and Hartwell, 1990). In this study, we use the competition mating assay to show that the a cell
can also make a choice of mating partner from among adjacent a cells. How do cells communicate to choose one from among several possible mating partners? A strong circumstantial case can be made for a central role of mating pheromones in this process. Although pheromoneless a cells (Kurjan, 1985) and pheromoneless a cells (Michaelis and Herskowitz, 1988) could be stimulated to undergo a mating response, these cells were not chosen as mating partners by cells of the opposite mating type, even in the presence of pheromone-producing wild-type a and a cells, respectively. Hence, even though both a- and a-pheromones are diffusible, pheromone produced by one cell is insufficient to allow mating of an adjacent nonproducing cell, and mating is restricted to those cells that produce pheromone. The authors suggested that a cell must present its own pheromone in order to be chosen as a mating partner by a cell of opposite mating type. This requirement for presentation may indicate that pheromone is the signal for partner selection, or that it is necessary for some later step in mating. The results of Bender and Sprague (1989) suggest the former is true for a cells, since they showed that a haploid cell need only express a-pheromone receptor and a-pheromone in order to mate with an a cell. However, the mating efficiency observed in these experiments was at least 5-fold lower than that of wild-type a and a cells, leaving open the possibility that some other signal is necessary for efficient choice of mating partner. We have examined the role of pheromones in partner selection. We showed previously that an a cell chooses for a mating partner the adjacent cell producing the highest level of a-pheromone (Jackson and Hartwell, 1990). In the present study, we show that an a cell chooses the adjacent a cell producing the highest level of a-pheromone. These results indicate that the mating hormones, a- and a-pheromone, are the signals for partner selection and that both cell types act as responders (by responding to the relative concentrations of complementary pheromone) and signalers (by secreting the pheromone signal) during courtship. We suggest that courtship results in the polarized morphogenesis of the responding cell in response to a gradient of pheromone in its environment. Because of the powerful genetic, biochemical, and cell biological techniques that can be applied to S. cerevisiae, studies of courtship in S. cerevisiae should provide insights into the mechanisms by which other eukaryotic cells generate cellular asymmetry in response to environmental signals. Pertinent examples are the polarized growth of nerve cells stimulated by nerve growth factor and other molecules (Dodd and Jessell, 1988), migrations of cells and axons in Caenorhabditis elegans (Hedgecock et al., 1990), migration of T-lymphocytes and their interaction with target cells during an immune response (Young and Cohn, 1988; Springer, 1990), and chemotaxis in Dictyostelium discoidium and polymorphonuclear leukocytes (Devreotes and Zigmond, 1988; Devreotes, 1989).
Cell 1040
COMPETITION
A
MATING ASSAY
Non-competition
Competition
Good competition
Index:
B
DISCRIMINATION
Discrimination
Randomness
0
ASSAY Random Choice
Index:
1
Figure 1. Schematic Diagrams of the Competition Mating and Discrimination Assays Used to Assess the Ability of Cells to Choose Mating Partners during Courtship In both assays, the ratio of responder x cells (represented by open circles) to signaling y cells (solid or stippled circles) is low so that each x cell is surrounded by y cells. The arrows represent potential interactions between the responder x and signaling y cells during courtship. (A) In the competition mating assay, only mating between an x and a target y cell (yr, solid circles) is measured; the ability of the challenger y cells (yc, stippled circles) to compete-that is, to inhibit mating between an x and a y, cell by attracting the x cell-is monitored. The ability to compete is quantified by computing the Cl (see Experimental Procedures). A noncompetitive ys cell fails to attract the responder x cell (resulting in a low Cl), whereas a yc cell that is a good competitor attracts an x cell as efficiently as a yc cell does (resulting in a Cl of 1). If a good cumpetitor yc strain is mating proficient, a zygote is formed between the responder x and yc cells; if not, the x and yc cells undergo the courtship interaction, but then fail to complete the process of diploid formation. (B) In the discrimination assay, the relative mating efficiency of the responder x cells with the two different types of signaling y cells is monitored. The signaling y cells produce different levels of pheromone, the courtship signal. Cells producing the higher level of pheromone are solid circles, and cells producing the lower level (usually no pheromone) are stippled. The two most extreme cases are shown: a responder x cell that discriminates perfectly chooses exclusively a cell producing the higher level of pheromone as a mating partner (re(,:esented by a randomness index of 0, see text), whereas a cell completely defective in discriminating the two types of y cells chooses a mating partner at random (represented by a randomness index of 1).
assay and the discrimination assay (Figure 1). in both assays, one ceil type (MATx or x) is given the choice of two different mating partners of opposite mating type (MA? or y). Since each assay monitors the ability of an x cell to choose one of the two types of y ceil as a mating partner, the x ceil is referred to as the responder and the y ceil the signaler in mating partner selection. In both assays, the ratio of x to y cells is low so that each x ceil is surrounded by y ceils. Hence, the responding role of the x cell can be assessed independently of the response of the y cells, since ail of the y cells surrounding a given x cell will respond only to that x ceil and therefore have no choice to make. In the competition mating assay, the two types of signaling y ceils are termed target and challenger. Diploids produced between x and only one of the potential y partners (the target ceil) are monitored. The role of the challenger ceil in producing the signal for mating partner selection is assessed by the effect it has on mating of the x with the target y cells. in this assay the x cell and the target y cell are normally wild type and the genotype of the challenger is varied to assess the role of various genes in signal production. The advantage of this assay is that the challenger does not need to be mating proficient to have a detectable effect. That some challenger cells do compete with a wild-type target y cell for mating with a responding x ceil while others do not establishes that a responder ceil can make a choice between potential partners. The finding that some mating-defective mutants can compete demonstrates that partner selection is distinct from and a prerequisite to conjugation (Jackson and Hartwell, 1990; see below). In the discrimination assay, the relative efficiency of mating of x ceils with the two potential y partners is measured. This assay is used to monitor the ability of the responding x ceil to discriminate between potential y partners. One y partner normally produces the signal for mating partner selection and the other does not (or produces a reduced amount). The advantage of this assay is that small defects in the ability of the responding cell to discriminate mating partners can be measured sensitively, since mating with one or the other y strain can be measured over a 104-fold range. To draw conclusions about the ability of the responding ceil to choose partners, it is necessary to show that the two y cells have the same intrinsic capacity to mate, independent of their ability to produce signal. This result is established by showing that there are conditions under which the poorer signaler is able to mate as efficiently as a wild-type strain. In several experiments, the concentration of a-pheromone by a cells is varied by using strains defective in one or both of the a-pheromone structural genes, MFa7 and MFa2. An mfal strain produces approximately 20-fold less a-pheromone not shown),
than
a wild-type mfa2
a strain
(Kurjan,
strain produces
1985;
data
Results
and an mfal omone (Kurjan, 1985).
no a-pher-
Experimentsi Design We use two assays to reach conclusions about interceiiuiar communication during mating: the competition mating
a Ceils Can Choose Among Potential a Mating Partners We showed previously that the a cell plays the role of re-
Mating 1041
Table
Partner
Selection
1. Competition
in S. cerevisiae
Mating
Assay
with Wild-Type
a
Responding
Theoretical a, Strait?
at Strains
rc
MFal MFa2 mfal mfa2 ste3 sfe3
MFa7 MFa2 MFal MFa2 MFal MFa2 mfal MFa2
0.22 0.24 0.24 0.25
f f + f
0.037 0.015 0.039 0.016
(4) (5) (3) (3)
Cells Cld for CI~ a
Noncompetitor
Good
0.24 0.25 0.25 0.27
1 1 1 1
f 0.032 f 0.016 f 0.034 f 0.014
Competitor
Observed 1.17 0.29 0.73 2.26
2 f f f
Mating
Cle 0.33 0.03 0.34 0.90
(4) (5) (3) (3)
Efficiency
24.2 f
