MOLECULAR AND CELLULAR BIOLOGY, JUlY 1991, p. 3603-3612 Vol. 11, No.

7
0270-7306/91/073603-12$02.00/0
Copyright © 1991, American Society for Microbiology

Significance of C-Terminal Cysteine Modifications to the Biological

Activity of the Saccharomyces cerevisiae a-Factor
Mating Pheromone
STEVAN MARCUS,1 GUY A. CALDWELL,1 DAVID MILLER,1 CHU-BIAO XUE,2
FRED NAIDER,2 AND JEFFREY M. BECKER'*
Department of Microbiology and Program in Cellular, Molecular, and Developmental Biology, University of Tennessee,
Knoxville, Tennessee 37996,1 and Department of Chemistry, College of Staten Island, City University of New York,
Staten Island, New York 103012
Received 9 October 1990/Accepted 22 April 1991
We have undertaken total synthesis of the Saccharomyces cerevisiae a-factor (NH2-YIIKGVFWDPAC[S-
farnesyl]-COOCH3) and several Cys-12 analogs to determine the significance of S-farnesylation and carboxy-
terminal methyl esterification to the biological activity of this lipopeptide mating pheromone. Replacement of
either the farnesyl group or the carboxy-terminal methyl ester by a hydrogen atom resulted in marked
reduction but not total loss of bioactivity as measured by a variety of assays. Moreover, both the farnesyl and
methyl ester groups could be replaced by other substituents to produce biologically active analogs. The
bioactivity of a-factor decreased as the number of prenyl units on the cysteine sulfur decreased from three to
one, and an a-factor analog having the S-farnesyl group replaced by an S-hexadecanyl group was more active
than an S-methyl a-factor analog. Thus, with two types of modifications, a-factor activity increased as the S-
alkyl group became bulkier and more hydrophobic. MATa cells having deletions of the a-factor structural genes
(mfal mfa2 mutants) were capable of mating with either sst2 or wild-type MATa cells in the presence of
exogenous a-factor, indicating that it is not absolutely essential for MATa cells to actively produce a-factor in
order to mate. Various a-factor analogs were found to partially restore mating to these strains as well, and their
relative activities in the mating restoration assay were similar to their activities in the other assays used in this
study. Mating was not restored by addition of exogenous a-factor to a cross of a wild-type MATa strain and a
MATaste6 mutant, indicating a role of the STE6 gene product in mating in addition to its secretion of a-factor.

Numerous proteins are now known to be modified post- shown that C-terminal methyl esterification is not required
translationally by addition of prenyl groups. The consensus for secretion of a-factor (41). The purpose of this work was
sequence for prenylation appears to be the C-terminal -CA to further study the significance of prenylation and methyl
AX box, where C is cysteine, A is an aliphatic amino acid, esterification to the biological activity of the S. cerevisiae
and X is any amino acid (18, 19); the modification involves a-factor.
addition of the prenyl group to the cysteine sulfur, followed Conjugation between S. cerevisiae a (MATa) and a
by proteolysis of the C-terminal -AAX (39, 55). Examples of (MATa) haploid cells to form a/a diploids requires the
prenylated proteins include yeast and mammalian RAS (6, reciprocal action of a-factor, produced by MATa cells, and
54, 55), the y subunit of yeast and mammalian G-proteins a-factor, a peptide pheromone secreted by MATa cells (12,
(16, 38), nuclear lamins (61), and various fungal mating 25, 26). These pheromones induce in their target cells similar
pheromones (2, 31, 51, 52). In some cases, but not all, the responses which are required for mating. Growth arrest
carboxyl group of the C-terminal cysteine is also methyl occurs at G1 of the cell cycle (5, 23), cell surface agglutinins
esterified (2, 11, 13). Recent studies (14, 37, 50) suggest that are produced (15), transcription of several genes is induced
other mammalian proteins are prenylated, indicating that (1, 33, 60), and the cells undergo a morphological alteration
this type of posttranslational modification may be common termed shmooing (36, 57). a-Factor, a tridecapeptide, is
to many proteins in eukaryotic organisms. processed via the well-characterized yeast secretory path-
In efforts to elucidate the biological significance of pren- way, and considerable studies on structure-activity relation-
ylation, it has been shown that this modification is required ships have been published (47). On the other hand, structure-
for proper localization and function of mammalian and yeast activity relationship studies on a-factor, which is apparently
RAS (48, 55) and for maximal biological activity of several secreted by a novel secretory pathway which has not been
fungal mating pheromones (2, 59, 63). In addition, prenyla- fully characterized (4, 28, 34, 39, 41, 43, 54, 55), have only
tion of the Saccharomyces cerevisiae a-factor seems to be recently been initiated. Anderegg et al. (2) found that re-
required for secretion of this peptide mating pheromone (48, placement of the famesyl group of a-factor by a methyl
54). The significance of the second modification, the car- group (S-methyl a-factor) caused only a moderate decrease
boxy-terminal methyl ester, found on some prenylated pro- in biological activity. These workers concluded that the
teins is not as well understood. It is known, however, that
this modification is important for the biological activity of famesyl group of a-factor was not required for biological
some fungal pheromones (2, 17, 63). In addition, it has been activity. We have previously shown that replacement of the
famesyl group of a-factor by a hydrogen atom (nonfarnesyl-
ated a-factor) results in an approximately 800-fold decrease
in biological activity (63).
*
Corresponding author. While a-factor in the absence of MATa cells is capable of
3603
3604 MARCUS ET AL. MOL. CELL. BIOL.

inducing a biological response in MATa cells, it had been Shmoo assays. RC757 and X2180-1B were grown at 30°C
suggested that MATa cells must actively produce a-factor in with shaking to early log phase in YEPD-BSA (1% yeast
order to mate (44). In those studies, MATa cells having extract, 2% peptone, 2% dextrose with bovine serum albu-
deletions of the a-factor structural genes (mfal mfa2 mu- min [BSA, 0.2 mg/ml] in 0.05 M citrate, pH 4.5). The cells
tants) were not capable of mating with wild-type (SST2+) were harvested by centrifugation at 1,000 x g, washed twice
MATa cells, even in the presence of exogenous a-factor. with sterile distilled water, and resuspended to 4 x 106 cells
This result led the authors to hypothesize that the a-factor per ml in YEPD-BSA. Dilution series (1:2) of a-factor and
found extracellularly is not sufficient for mating but rather a-factor analogs were prepared in siliconized borosilicate
that a MATa cell-bound form of a-factor might be required glass tubes with YEPD-BSA as the diluent. Five hundred
for mating. Recent studies by Jackson and Hartwell further microliters of cell suspension was added to 500 ,ul of each of
demonstrated the importance of both a- and a-factor produc- the a-factor and a-factor analog dilutions, and the resulting
tion in a process they termed courtship (29, 30), a process by cell suspensions were incubated for 3.5 h at 30°C with
which opposite mating types select the mating partner which shaking. At the completion of the incubation period, the cells
produces the most pheromone. were placed on ice, and 10-pul portions of each suspension
To further study a-factor structure-activity relationships, were placed in a hemacytometer and observed microscopi-
we synthesized several a-factor analogs and examined the cally to quantitate the total number of cells, percent shmoos
role of C-terminal cysteine modifications in the biological (elongated, pear-shaped cells [36, 57]), and percent unbud-
activity of this lipopeptide mating pheromone. In this report, ded cells. The endpoint of activity in this assay was defined
we show by use of four different and sensitive biological as the concentration of pheromone or pheromone analog
assays that removal of either the farnesyl group or the which caused approximately 50% of the cells to shmoo. At
carboxy-terminal methyl ester of a-factor leads to a similar higher concentrations of pheromone, 100% of treated cells
reduction in the bioactivity of the modified pheromones. In were shmoos. The profiles of the response curves to all
addition, we show that the farnesyl group is not specifically analogs were highly similar, so that the amounts of phero-
required for high biological activity, since replacement of mone causing 10% shmoos, 50% shmoos, and 100% shmoos
this group with any of several different hydrophobic hydro- were nearly identical for all pheromones tested.
carbon moieties resulted in analogs with biological activity Halo assay. YEPD (YEPD [pH 4.5] with 2% Bitek agar
close to that of a-factor. We further show that addition of [Difco]) plates were overlaid with 4 ml of RC757 cells (2.5 x
exogenous a-factor allows MATa mfal mfa2 mutants to mate 105 cells per ml) in Noble agar (0.825%). Five-microliter
with sst2 or wild-type (SST2+) MATa cells, demonstrating portions of a-factor solutions at various concentrations in
that it is not absolutely essential for MATa cells to actively BSA (0.2 mg/ml of water) were then spotted onto the
produce a-factor in order to mate. overlay. The plates were incubated at 30°C for 24 to 36 h and
then observed for clear zones (halos), an indication of the
MATERIALS AND METHODS arrest of cell growth, at the various sample locations (Fig. 2).
FUSI-lacZ induction assay. S. cerevisiae LM23-116az was
Strains. Halo and shmoo assays were carried out with grown overnight in YEPD at 30°C. Cells were diluted to a
wild-type S. cerevisiae X2180-1B (MATa) (Yeast Genetics concentration of 5 x 106/ml and grown to 1 x 107 to 2 x
Stock Center) or a mutant supersensitive to a-factor, RC757 107/ml at 30°C. Cells were concentrated by centrifugation
(MATa sst2-1 rmel his6 meti can] cyh2) from Russell Chan and then resuspended at 108/ml. Induction was performed by
(7, 8). FUSI-lacZ assays were carried out with S. cerevisiae adding 0.5 ml of a-factor, a-factor analog, or a-factor dilu-
LM23-116az [MATt FUSI: :lacZ leu2 lys5 meti ura3-52 tions or water to 4.5 ml of concentrated cells. These mix-
ste2::TRPI] from Lorraine Marsh. Mating experiments were tures were vortexed and placed at 30°C with shaking for 2 h.
done with S. cerevisiae RC757 (above), CR11 (MATa After this time, cells were harvested in a tabletop centrifuge,
cryl-11 thr4 his6 lysi) from Wolfgang Duntze (56), SM1058 and each pellet was resuspended in 0.5 ml of Z buffer (16.1 g
(MATa trpl his4 leu2 ura3 cani) and SM1229 (isogenic to of Na2HPO4 7H20, 5.5 g of NaH2PO4. H20, 0.75 g of KCl,
SM1058 but mfalA::LEU2 mfa2A::URA3) from Susan Mi- 0.246 g of MgSO4. H20, 2.7 ml of ,-mercaptoethanol per
chaelis (44), and JPY200 [MATa gal2 his3A200 leu2-3,112 1,000 ml). For each sample, the OD6. was recorded, and 1:5
lys2-801 trp-l ura3-52] and JPY201 (isogenic to JPY200 but and 1:10 dilutions of cells in Z buffer in a total volume of 1.0
ste6A-1 :HIS3) from John McGrath (43). ml were prepared. Yeast cells were permeabilized by the
Synthesis and characterization of a-factor and Cys-12 ana- addition of 75 p1l of 0.1% sodium dodecyl sulfate and 60 pl of
logs. The a-factor (NH2-YIIKGVFWDPAC[S-farnesyl]- chloroform, followed by vortexing for 15 s. After equilibra-
COOCH3) and a-factor analogs used in this study were tion to 30°C, the mixtures were assayed for ,-galactosidase
synthesized either by solution-phase peptide synthesis or by activity. Assays were initiated by the addition of 0.2 ml of
a combination of solid-phase and solution-phase peptide O-nitrophenyl-,-D-galactopyranoside (4 mg/ml in 0.1 M
syntheses; the structures of all analogs are shown in Fig. 1. KPO4 [pH 7]) to each sample. After the appearance of a
Detailed protocols for the synthesis and characterization of medium yellow color (10 min to 4 h), reactions were stopped
a-factor, nonfarnesylated a-factor, and nonmethylated a-fac- by adding 0.5 ml of 1 M Na2CO3. Cells from all assay
tor have been published previously (62, 63). Details of the mixtures were pelleted, the supernatant fractions were read
synthesis and characterization of all other analogs will be at OD420 and OD550, and units of P-galactosidase activity
published separately. All peptides were purified to greater were calculated (21, 45, 49). The endpoint of activity in this
than 98% homogeneity by reversed-phase high-pressure assay was defined as the concentration of pheromone re-
liquid chromatography (HPLC) and were characterized by quired to induce a response in ,-galactosidase units that was
amino acid analysis, 400 MHz 1H nuclear magnetic reso- twofold greater than the basal level observed in either
nance spectroscopy, and fast atom bombardment mass spec- untreated LM23-116az cells or LM23-116az cells treated
troscopy. a-Factor and a-factor analogs were dissolved with a-factor (1 ,ug/ml). Endpoints were calculated as the
directly at 1 to 5 mg/ml in methanol for use as stock solutions average value from three separate trials, with each trial being
in all biological assays. performed in duplicate. In all FUSI-lacZ assays performed,
VOL . 1 l, 1991 a-FACTOR STRUCTURE-ACTIVITY RELATIONSHIPS 3605

-Factor:
Tyr4.4mLy.GIy.vaI.Ph...Asp4PrACys-OCH3

C3 e 3 <3

Removal of famesyl and/r carboxymethyl group Alteration of he Cys-thlotwgroLp

Unmodified dodeapeptide: -Cys-OH S9thyl A-factor: 43OCH3
IH CH3

Nonm.thylated p-fetor: -Cys-OH S-Hexadecanyl n-factor: s-OCH3

C%3 H3

Nonfamesylated I-factor: s-OCH3 S-Benyl a-factor: -C-OCH3

H
tH2

Alteration of he carboxy/ modification

S-Prenyl -factor:
Cys-anmdd a-factor: -Cys-NH2

CH3
-
3 C3 C3
S-Geranyl n-factor: -C s-OCH3
3-methyl butyl
n-factor:
estr s-OCH2H2CH(CH3)2
H
CH3 iH3 H3

3-mothyl-2butenyl ester S-OCHjHc(H3)2

n-factor:
CH3 CH3 H3

FIG. 1. Structures of S. cerevisiae a-factor and various Cys-12 analogs used in this study. All analogs are denoted only by the Cys
modification in this figure. All of the analogs maintained the full peptide backbone of YIIKGVFWDPAC.

the level of ,-galactosidase induction correlated to the at room temperature for about 1 h prior to spotting of
concentration of pheromone used, so that a gradual increase a-factor. a-Factor and a-factor analog dilutions were pre-
in ,B-galactosidase units resulted from a corresponding in- pared with BSA (2 mg/ml) as the diluent, and each dilution
crease in the concentration of pheromone tested. was spotted onto the mixed lawn of cells. The plates were
Mating restoration assay. The mating restoration assay incubated at 30°C for 2 days, after which constellations of
used in this study was a modification of the procedure diploid colonies were apparent at the positions where a-fac-
developed by Horecka and Sprague (27). For dose-response tor was spotted onto the mixed lawn (Fig. 3). The endpoint
assays, S. cerevisiae RC757 and SM1229 were grown to late of a-factor activity was defined as the lowest amount of
log phase in YEPD at 30°C with vigorous shaking. Cells were a-factor which produced a constellation of diploid colonies.
pelleted by centrifugation, washed two times with cold Mating efficiency assays were performed by a modification
YEPD, and then resuspended in cold YEPD at 107/ml. Equal of the dose-response assay discussed above and the mating
volumes of the two suspensions were mixed together, and assay used by Michaelis and Herskowitz (44). S. cerevisiae
this mixed cell suspension was diluted 1:3 with sterile MATa and MATa strains were grown to late log phase in
distilled water. A 900-,ul amount of the diluted mixed sus- YEPD at 30°C with vigorous shaking. Cells were pelleted by
pension was plated onto SD medium (yeast nitrogen base centrifugation and then washed two times with cold YEPD.
without amino acids [Difco], 2% dextrose, and 2% Bitek agar MATa cells were resuspended in cold YEPD at 108/ml.
[Difco]) for selection of diploids. The plates were incubated Wild-type (SM1058 and JPY200) MATa cells were resus-
3606 MARCUS ET AL. MOL. CELL. BIOL.

TABLE 1. Biological significance of the carboxy-terminal methyl

ester and farnesyl groups of a-factor
% of wild-type activitya
a-Factor or analog Halo Shmoo FUSI-
assay assayb lacZ assay
a-Factor (wild type) 100 100 100
Nonmethylated a-factor 1.5 0.8 0.05
Nonfarnesylated a-factor 0.1 0.1 0.1
S-Methyl a-factor 0.2 0.4 NTc
Unmodified a-factor peptide 0.0025 NT NT
a
Wild-type a-factor activity in each assay was as follows: halo assay, 0.06
ng; shmoo assay, 0.04 ng/ml; FUSI-lacZ assay, 0.03 ng/ml. Each value
represents an average of at least three determinations, with a variation of not
FIG. 2. Halo assay of synthetic a-factor. A lawn of RC757 MATTa more than twofold for each value reported.
b The tester strain used in this shmoo assay was RC757 (sst2).
cells was prepared as described in Materials and Methods. A 1:2
dilution series of synthetic a-factor was prepared in BSA (0.5 c NT, not tested.
mg/ml), and 5 ,ul of each dilution was spotted on the lawn. Phero-
mone bioactivity resulted in the formation of clear zones, or halos,
at the areas where samples were spotted, an indication of growth
arrest. X2180-1A (MATa) and X2180-1B (MATa) were patched, and
diploids per mm2. If a total of 1.5 x 106 MATa cells (from 900
5 ,ul of concentrated MATa culture medium was spotted onto the ,ul of a 1:3 dilution of the mixture of SM1229 and RC757
lawn, as indicated, all as controls. As indicated, the endpoint of cells) were plated on an 85-mm-diameter agar plate, the
a-factor activity was 0.06 ng. density of MATa cells was 264 cells per mm2. The mating
efficiency was then calculated as the observed colony den-
sity in the diploid constellation (5 diploids per mm2) divided
pended in cold YEPD at concentrations of 105/ml for deter- by the density of MATa cells (264 potential diploids per
mination of normal wild-type mating efficiency and 108/ml mm2), which gives a value of 1.9 x 10-2. In some cases, a
for determination of mating efficiency in the presence of dissecting microscope was used to visualize accurately the
exogenous a-factor. mfal mfa2 (SM1229) and ste6 (JPY201) individual colonies in a small area. Under the conditions
MATa strains were resuspended in cold YEPD at 107/ml for described (1.5 x 107 MATot cells per 85-mm-diameter petri
crosses with RC757 and 108/ml for crosses with CR11. Equal plate [2,643 cells per mm2]), wild-type strains (CR11 x
volumes of MATa and MATa suspensions were mixed SM1058) mated with an average efficiency of about 20%, as
together, and then the mixed cell suspensions were diluted determined in three independent trials. However, mating
1:3 with sterile distilled water. The remainder of the assay efficiency increased to about 60% when 5 x 107 MATot cells
was the same as that used for the dose-response assays. The per plate were used (8,811 cells per mm2). In order to
efficiency of mating was calculated as the ratio of the number compare and calibrate the above-outlined mating assay to
of diploid colonies formed at the affected area (constellation the filter mating assay (24) used by some other investigators,
of diploid colonies) to the total number of MATa cells we also crossed CR11 and SM1058 to measure the mating
calculated to be present in that area (as calculated from the efficiency by that method. The mating efficiency of these
total number of MATa cells plated). For example, if the strains in the filter mating assay was 24%, which correlated
diameter of a constellation of diploids was 10 mm and it closely with our values from the mating restoration assay
contained 400 colonies, then the diploid density was 5 used in this study.

RESULTS
Biological significance of the carboxy-terminal methyl ester
and farnesyl groups of a-factor. For our analyses of the
biological activity of a-factor and Cys-12 analogs, we used
four different assays: the halo assay, an agar diffusion assay
which measures pheromone response based on growth arrest
of a cells (Fig. 2); the shmoo assay, an assay which measures
pheromone response by morphological alteration (shmooing)
and growth arrest of a cells; the FUSI-lacZ induction assay,
an assay which measures pheromone response by induction
of a FUSJ-lacZ reporter gene (42, 58); and the mating
restoration assay, which is discussed in more detail below.
MA Ta mfa 1 mfa2and MATa sst2 I To address the significance of the farnesyl and methyl ester
modifications to the biological activity of a-factor, we tested
FIG. 3. Mating assay with synthetic a-factor. A mixed lawn of a-factor, unmodified a-factor dodecapeptide, nonfamesyl-
RC757 and SM1229 was prepared on minimal medium for selection ated a-factor, and nonmethylated a-factor (Fig. 1). As shown
of diploids as described in Materials and Methods. A 1:2 dilution in Table 1, for three different and sensitive bioassays, our
series of synthetic a-factor was prepared in BSA (0.5 mg/ml), and 5
,ul of each dilution was spotted onto the mixed lawn of cells. The results clearly indicate that removal of either the farnesyl or
plates were incubated at 30°C for 2 to 3 days, after which constel- methyl ester group leads to approximately a 100- to 1,000-
lations of diploid colonies were apparent at the positions where fold reduction in bioactivity. The unmodified a-factor dode-
a-factor was spotted onto the mixed lawn. As indicated, the end- capeptide was virtually inactive, showing an endpoint of
point of a-factor activity was 0.06 ng. activity of 2.5 p.g in the halo assay (40,000-fold less active
VOL. 11, 1991 a-FACTOR STRUCTURE-ACTIVITY RELATIONSHIPS 3607

TABLE 2. Activity of a-factor and various analogs in the shmoo TABLE 4. Biological activity of a-factor analogs with altered
assay with SST2+ MATa cells C-terminal carboxyl groupsa
a-Factor or analog o% ofactivity'
wild-type % of wild-type activity
a-Factor or analog
Halo assay Shmoo assay
a-Factor (wild type) ........................................ 100
Nonmethylated a-factor ........................................ 6.5 a-Factor (wild type) 100 100
Nonfarnesylated a-factor ...................................... 1.6 Nonmethylated a-factor 1.5 0.8
S-Methyl a-factor .............................. .......... 6.5 Cys-amide a-factor 6 6
Unmodified a-factor peptide .................................. NAb 3-Methyl butyl ester a-factor 0.8 0.4
a The SS2+ strain used in this shmoo assay was X2180-1B. The assay was 3-Methyl-2-butenyl ester a-factor 1.5 0.8
carried out as described in Materials and Methods except that YNB was used a See Table 1, footnotes a and b.
as the assay medium. Wild-type a-factor activity in this assay was 160 ng/ml.
Each value represents an average of at least three determinations, with a
variation of not more than twofold for each value reported.
b NA, not active at 40 ,ug/ml.
of the farnesyl group with a hexadecanyl group produced an
analog that was 30- to 70-fold less active in the shmoo and
halo assays and 7-fold less active in the FUSI-lacZ assay.
than a-factor). The endpoint for a-factor response, as defined When the sulfur group was modified with an S-benzyl
in Materials and Methods, was approximately the same in moiety, the resulting analog exhibited activity that was 4- to
the shmoo assay (40 pg/ml) and the FUSI-lacZ assay (30 30-fold lower than that of a-factor in the shmoo and halo
pg/ml). The rank order of activities of the various analogs assays. The effect of alteration of the length of the Cys-
was approximately the same for all assays. isoprenyl group was determined by analysis of the activity of
Our results suggest that the farnesyl group and carboxy- S-prenyl a-factor and S-geranyl a-factor. S-Prenyl a-factor
terminal methyl ester play nearly equal roles in the biological was found to be 15- to 70-fold less active than a-factor, while
activity of a-factor. This conclusion is in contrast to that S-geranyl a-factor was only 2- to 8-fold less active. Indeed,
made by Anderegg et al. (2), who concluded that only the the S-geranyl a-factor is the most potent synthetic analog
carboxy-terminal methyl ester was required for activity. prepared to date.
However, it should be pointed out that in their study, Consequences of altering the C-terminal carboxy modifica-
S-methyl a-factor, and not nonfarnesylated a-factor (S-H tion of a-factor. Having determined the relative significance
a-factor), was tested. For comparative purposes, we deter- of the Cys-thioether modification of a-factor, we focused our
mined that S-methyl a-factor was about two- to four-fold attention on the C-terminal methyl ester modification. The
more active than nonfarnesylated a-factor in the halo and following analogs were synthesized and analyzed: Cys-
shmoo assays (Table 1). In our assays, sst2 MATa cells were amide a-factor, 3-methyl butyl ester a-factor, and 3-methyl-
used, and the assay medium was YEPD. Anderegg et al. (2) 2-butenyl a-factor (Fig. 1). The Cys-amide a-factor was the
used SST2+ MATo cells (X2180-1B) and YNB as the assay most active of this series of analogs, being about 16-fold less
medium in their a-factor bioactivity assays. We tested the active than a-factor (Table 4). 3-Methyl-2-butenyl ester
analogs shown in Table 1 under the assay conditions of a-factor and 3-methyl butyl ester a-factor did not exhibit
Anderegg et al. (2) and obtained a rank order of activity activities above that observed for nonmethylated a-factor,
(Table 2) that was similar to the rank order obtained when each being about 100-fold less active than a-factor.
sst2 MATa cells and YEPD were used (Table 1). MATa cells need not actively produce a-factor in order to
Consequences of altering the size and hydrophobicity of the mate with either sst2 or wild-type MATa cells. To investigate
Cys-thioether group of a-factor. To determine whether the further the biological significance of extracellular a-factor,
farnesyl group of a-factor is specifically required for full we tested a-factor and several Cys-12 analogs in a recently
bioactivity, we synthesized the following a-factor analogs: developed mating restoration assay (27). In this assay, a
S-methyl a-factor, S-hexadecanyl a-factor, S-prenyl a-fac- mixed lawn of MATa cells defective in a-factor production
tor, S-geranyl a-factor, and S-benzyl a-factor (Fig. 1). Re- (mfal mfa2 mutants) and MATa cells were prepared on
placement of the famesyl moiety by a methyl group (S- medium selective for growth of diploid cells. In the absence
methyl a-factor) produced an analog which was about 200- to of exogenously added a-factor, mfal mfa2 mutants could not
500-fold less active than a-factor and 2- to 4-fold more active mate with wild-type MATa cells and mated with sst2 MATa
than nonfarnesylated a-factor (Tables 2 and 3). Replacement cells at very low efficiency (0.038% of wild-type mating
level) (Table 5). However, when a-factor was spotted on the
lawn, constellations of diploid colonies grew at the locations
TABLE 3. Biological activity of a-factor analogs with altered where the pheromone was spotted (Fig. 3). Although mating
Cys-thioether groups' was not restored to a wild-type level (Table 5), the results
% of wild-type activity demonstrate that it is not absolutely essential for MATa cells
a-Factor or analog to actively produce a-factor in order to mate with wild-type
Halo Shmoo FUSI-lacZ MATa cells. The endpoint for mating restoration with wild-
assay assayb assay type MATa cells was about the same as the endpoint
a-Factor (wild type) 100 100 100 obtained in the halo assay for the same cells (2 ,ug). As
Nonfarnesylated a-factor 0.1 0.1 0.1 shown in Table 6, the proficiency with which various a-fac-
S-Methyl a-factor 0.2 0.4 NTc tor analogs restored mating correlated, for the most part,
S-Hexadecanyl a-factor 1.5 3.3 15 with activities found in the halo assay. However, the S-pre-
S-Prenyl a-factor 1.5 6.7 NT nyl, S-geranyl, and S-benzyl a-factors exhibited activities
S-Geranyl a-factor 12 50 NT higher than those found in the halo assay.
S-Benzyl a-factor 3 25 NT
One possible explanation for the inability of exogenous
a See Table 1, footnotes a, b, and c. pheromone to restore mating of the mfal mfa2 mutant to a
3608 MARCUS ET AL. MOL. CELL. BIOL.

TABLE 5. Exogenous a-factor restores mating of the mfal mfa2 TABLE 7. Exogenous a-factor does not restore mating of the
mutant to either wild-type or sst2 MATa cells ste6 mutant to a wild-type level
MATa a-Factor % of wild-type MATa a-Factor % of wild-type
MATa strain straina added' mating efficiencyc MAToL strain straina addedb mating efficiencyc
CR11 (wild type) Wild type - 100 CR11 (wild type) Wild type - 100
mfal mfa2 - <1.0 x 10-6 ste6 - <1.0 x 10-6
Wild type + 9.5 x 10-2 Wild type + 2.1 x 10-1
mfal mfa2 + 6.8 x 10-2 ste6 + <1.0 x 10-6
RC757 (sst2) Wild type - 90 RC757 (sst2) Wild type - 90
mfal mfa2 - 3.8 x 10-2 ste6 - 4.1 X 10-2
Wild type + 9.0 x 10- Wild type + 9.3 x 10-1
mfal mfa2 + 4.3 ste6 + 7.1 x 10-1
a The MATa strains tested were SM1058 (wild type) and SM1229 (mfal a The MATa strains tested were JPY200 (wild type) and JPY201 (ste6).
mfa2).
I
"See Table 5, footnote b.
A + indicates that 4 ,ug of a-factor was spotted onto the lawn. c See Table 5, footnote c. The actual average mating efficiency of the
I
The mating restoration assay used for determination of mating efficiencies wild-type strains used in this assay was 25%.
is described in the Materials and Methods section. All values are expressed as
percent wild-type mating efficiency, which was normalized to lOo. The
actual average mating efficiency of the wild-type strains used in this assay was
21%, as determined in three independent trials. case. When a-factor was added to a mating mixture of
wild-type MATa and MATa cells, the mating efficiency was
decreased to nearly the same level as was observed when
exogenous a-factor was added to a cross of mfal mfa2
wild-type level is that a-factor plays an intracellular role that mutants and wild-type MATa cells. Similar decreases in
cannot be complemented by exogenous pheromone. Since mating efficiency were observed when sst2 MATa cells were
ste6 mutants accumulate a-factor intracellularly without used (Table 5).
secreting the pheromone (34), we tested the ability of a ste6
mutant to mate with either wild-type or sst2 MATa cells in
the mating restoration assay. Exogenous a-factor did not DISCUSSION
restore mating at all when the ste6 mutant was crossed with Prenylation is now recognized as an important posttrans-
wild-type MATa cells (Table 7). Some degree of restoration lational modification of numerous proteins in eukaryotic
was obtained when the ste6 mutant was crossed with sst2 organisms. In this article, we have presented the results of a
MATa cells, but the level of restoration was lower than that detailed study of structure-activity relationships for a preny-
obtained with the mfal mfja2 mutant. lated yeast mating pheromone, the S. cerevisiae a-factor.
A second possible reason for the low level of mating Previous studies on the prenylated peptide pheromones,
restoration by exogenous a-factor is that exogenous phero- rhodotorucine A and tremerogen A-10, of the heterobasidio-
mone interferes with mating partner selection by MATa mycetous yeasts Rhodosporidium toruloides and Tremella
cells. Jackson and Hartwell (29, 30) showed that S. cerevi- mesenterica, respectively, demonstrated that hydrophobic
siae MATa and MATa haploids preferentially select mating modification of the C-terminal cysteines of each of these
partners which produce the highest level of pheromone, mating pheromones was required for biological activity (17,
suggesting that the cells might respond to pheromone gradi- 59). Detailed studies on analogs of tremerogen A-10 demon-
ents. If this hypothesis is correct, then destruction of the strated that decreasing the number of prenyl units from three
gradient, as would occur after addition of high levels of to one resulted in a corresponding decrease in biological
exogenous pheromone, should markedly decrease the level activity (17). In the present study, we have shown that
of wild-type mating. As shown in Table 5, this is indeed the absence of either the farnesyl group (NH2-YIIKGVFWDPA
C-COOCH3) or carboxy-terminal methyl ester (NH2-YIIKG
VFWDPAC[S-farnesyl]-COOH) of a-factor results in a sim-
TABLE 6. Activity of a-factor and various analogs in the ilar marked decrease but not loss of biological activity, as
mating restoration assaya determined by four sensitive bioassays. In addition, an
% of wild-type analog lacking both the farnesyl group and methyl ester
a-Factor or analog activitya (NH2-YIIKGVFWDPAC-COOH) was virtually inactive, be-
a-Factor (wild type) ........................................ 100
ing 40,000-fold less active than a-factor. These results sup-
Nonmethylated a-factor ........................................ 1.5 port those in previous studies on the S. cerevisiae a-factor
Nonfamesylated a-factor ...................................... 0.8 which utilized only the halo assay (63) and suggest that both
S-Methyl a-factor ........................................ 1.5 the farnesyl group and carboxy-terminal methyl ester of
S-Hexadecanyl a-factor ........................................ 0.8 a-factor are important for high biological activity. This
S-Prenyl a-factor ............................. ........... 25 conclusion is in contrast to that of Anderegg et al. (2), who
S-Geranyl a-factor ........................................ 200 concluded that only the methyl ester modification of the
S-Benzyl a-factor ........................................ 25 a-factor peptide was required for high biological activity.
Cys-amide a-factor ........................................ 6 However, Anderegg et al. (2) examined an S-methyl a-factor
3-Methyl butyl ester a-factor ................................. 0.4 rather than S-H (nonfamesylated) a-factor. Furthermore,
3-Methyl-2-butenyl ester a-factor ........................... 0.8 they did not test concentrations of a-factor analogs more
a
The mating restoration assay is described in the Materials and Methods than fivefold higher than the concentration of a-factor tested.
section. RC757 (sst2) was used as the MATca strain. All values are expressed This prevented them from detecting activity from analogs
as percent wild-type activity. Wild-type a-factor activity in this mating that were significantly less active than a-factor.
restoration assay was 0.06 ng. Each value represents an average of at least
three determinations, with a variation of not more than twofold for each value Additional a-factor analogs were studied to determine
reported. whether S-alkyl modifications other than farnesyl would
VOL . 1 l, 1991 a-FACTOR STRUCTURE-ACTIVITY RELATIONSHIPS 3609

result in biologically active lipopeptides. We found that receptor of MATa cells, but also the ability of hydrophobi-
aliphatic, benzylic, and geranyl substituents resulted in cally modified peptides to enter the plasma membranes of
pheromone analogs exhibiting from 0.2 to 50o of the activity these cells. It is conceivable that the hydrophobic a-factor
of the natural mating factor (Table 3). As illustrated by the might enter the lipid bilayer of MATa cells and then diffuse
S-benzyl analog, an isoprenyl moiety is not required for high through the membrane to bind its receptor. A similar process
biological activity. The bioactivity of the pheromone in- has been proposed for the interaction of mammalian peptide
creased as the length of the prenyl moiety was increased hormones with the receptors of their target cells (3, 53).
(S-farnesyl a-factor > S-geranyl a-factor > S-prenyl a-fac- An additional factor which might have an effect on the
tor), and the S-hexadecanyl a-factor was more active than apparent activities of a-factor analogs is MATa cell degra-
the S-methyl analog. Thus, within these two types of S-alkyl dation of a-factor. Previous studies on rhodotorucine A, the
modifications, a-factor activity was found to increase as the pheromone produced by mating type A cells of R. toruloides,
S-alkyl group became bulkier and more hydrophobic. have demonstrated that proteolysis of this farnesylated
Replacement of the methyl ester by larger, more hydro- peptide by the mating type a target cell is a requirement for
phobic esters resulted in a marked decrease in activity induction of the pheromone response (46). In contrast, S.
(Table 4). In contrast, the Cys-amide a-factor (Fig. 1) had 6% cerevisiae MATa cells proteolyze a-factor as one mechanism
of the activity of the native pheromone. Thus, the ester of recovery from pheromone-induced growth arrest (10).
oxygen is not specifically required for interaction with the Mutants which are unable to proteolyze this mating peptide
a-factor receptor, and some hydrophobic groups are not well (sstl or bar] mutants) still respond and are, in fact, super-
tolerated in the ester moiety. We are presently synthesizing sensitive to a-factor (7, 8). These and other studies have
additional analogs to investigate this point in more detail. demonstrated that hydrolysis of a-factor by MATa cells is
Some differences in the activity of a-factor analogs relative not required for a pheromone response. We have recently
to a-factor were found, depending on the assay used for identified a MATa cell-specific a-factor-degrading activity
measuring activity. One must bear in mind that while each of (40). As is the case for a-factor proteolysis by MATa cells,
the a-factor analogs contains the same peptide moiety, they proteolysis of a-factor seems to be a mechanism of MATa
differ in their overall physical characteristics, especially with cell desensitization from the pheromone response. The sub-
regard to hydrophobicity, which influences peptide solubility strate specificity of the a-factor-degrading protease appears
in different media and susceptibility to degradation by a to depend on both the peptide sequence and the cysteine-
recently identified MATa cell-specific a-factor protease (40). modifying group. For example, S-hexadecanyl a-factor is not
Furthermore, the conditions of each assay, as well as the proteolyzed by MATa cells. We have not determined
specific physiological response measured, are different. whether all the analogs tested in this study are proteolyzed
Thus, the threshold for response and the stringency of by MATa cells, but the preferential degradation of a-factor
response in each assay could be expected to vary. Indeed, analogs may influence their relative bioactivities.
previous studies have demonstrated that various pheromone Because of its lipophilic nature and results showing that
responses require markedly different minimum doses of MATa mutants defective in a-factor production were inca-
a-factor, from 10-11 M for agglutination to 10-8 M for pable of mating with wild-type (SS12+) MATa cells even
shmooing (for review, see reference 12). Each of the four when a-factor was provided exogenously, some speculation
assays used in this study measures the end result of a had been raised that a-factor may be tethered to MATa cells
complex series of intracellular and/or intercellular events. and that its primary action on MATa cells occurs via cell-cell
While great strides have been made in characterizing specific contact (26, 44). However, since a-factor is found in the
aspects of the pheromone response pathway, it is on the culture supernatant of MATa cells, the pheromone is clearly
whole still poorly understood. However, there is some released or secreted by these cells. Moreover, as demon-
evidence that the biochemical events subsequent to phero- strated by numerous studies, the presence of MATa cells is
mone induction follow a branching pathway, with physiolog- not required for MATa cells to respond to a-factor supplied
ical responses as the end result of different branches (9, 32). as unpurified or partially purified preparations from spent
Therefore, some of the various responses measured by medium of MATa cells. Our results demonstrate that syn-
different bioassays may reflect the different thresholds for thetic a-factor can function to restore the ability of MATa
different branches of the pathway. mutants defective in a-factor production to mate with either
While it is clear from our analyses of the biological activity sst2 or wild-type MATa cells (Table 5), although mating was
of a-factor and various Cys-12 analogs that hydrophobic not restored to a wild-type level in either case. We found that
modification of the C-terminal cysteine of this pheromone is the relative rank order of activities of the various a-factor
required for high biological activity, it is equally evident that analogs tested in this mating restoration assay was similar to
activity is dependent on more subtle factors than just the the rank order of activities obtained in the halo assay (Table
overall hydrophobicity of the C terminus of a-factor. Char- 6), which is also done in agar. The mating restoration
acterization of the a-factor analogs by HPLC (data not experiments indicate that (i) MATa cells do not have to
shown) demonstrated that the S-hexadecanyl a-factor is the actively produce a-factor in order to mate with MATa cells,
most hydrophobic analog and that both the 3-methyl butyl although a-factor production might be required for mating at
and the 3-methyl-2-butenyl ester analogs are more hydro- a wild-type level and (ii) the cysteine modifications of
phobic than the a-factor. It is possible that optimal interac- a-factor are no more or less significant for inducing a MATa
tion of the pheromone with the receptor requires a long cell response to pheromone than for allowing MATa and
branched hydrocarbon on the cysteine sulfur and a smaller MATa cells to conjugate. A possible explanation for the
organic group on the cysteine carboxyl. Such a requirement inability of concentrated MATa culture supernatant fractions
would be similar to those models used to explain the to restore mating of mfal mfa2 mutants to wild-type MATa
interaction of aspartamelike sweeteners with their receptor cells in an earlier study (44) is that the preparation did not
(20). The possibility must also be considered that the re- contain a sufficient amount of a-factor to restore mating or
quirement of hydrophobic modification of a-factor for bio- contained a substance that inhibited mating.
logical activity reflects not only the specificity of the a-factor As indicated above, exogenous a-factor could not restore
3610 MARCUS ET AL.

|
u

.,. ...
.hR,-; . ':t-\.
F|Phrmo ne s secreted by both MA Ta and tfA Ta ceI Is
a-factor gradient

..>
... .............

usa^.:

. ..
....... .. ..
..

..........

........
a-factor gradient
B
ax-Factor secretion and a-factor supplementation

.
a-factor gradient

.......
:
............
........
....
no ra-actorgra dien t
MOL. CELL. BIOL.

FIG. 4. Model to explain why exogenous a-factor does not restore mating of mfal mfa2 mutants to a wild-type level. (A) Under normal
conditions, apposed MATa and MATa cells produce a-factor and a-factor, respectively, resulting in pheromone gradients which are highest
in concentration at the surface of the secreting cells. Shmoo tip projections are formed in the direction of the gradients, resulting in successful
courtship. (B) When an mfal mfa2 mutant is in apposition to a wild-type MATTa cell, the mfal mfa2 mutant is unable to court the MATa cell.
Instead, since a-factor is supplied exogenously, the MATa cell forms its projection randomly.

the mating efficiency of mfal mfa2 mutants to a wild-type courtship response of S. cerevisiae MATa and MATa hap-
level (Table 5). However, it is important to note that in loids. These investigators showed that when MATa and
previous studies (35), addition of exogenous a-factor also did MATa cells are given a choice of several different potential
not restore mating to wild-type levels in mfal mfa2 MATa mating partners, the cell type(s) which produces the most
cells. A possible explanation for the inability of exogenous pheromone is preferentially selected. The authors concluded
pheromones to restore mating of pheromone-defective that mating partner selection occurs via a response to
MATa or MATa cells to a wild-type level is that a- and pheromone gradients. Our results are, in fact, predictable
a-factor play an intracellular role that cannot be comple- from Jackson and Hartwell's model for courtship (29, 30).
mented by exogenous pheromone. To examine this possibil- Normal courtship occurs when MATa and MATa cells are
ity, we tested a ste6 mutant in the mating restoration assay. each producing normal levels of pheromone, which results in
The STE6 gene product has a high degree of homology to recognizable gradients of the respective pheromones toward
mammalian multidrug resistance transporters (34, 43) and is which projections are directed (Fig. 4A). However, when the
thought to act as a transporter which secretes a-factor. No MATa cell does not produce a-factor (mfal mfa2 mutant) and
other role for STE6 has been proposed, and it has been the pheromone is instead provided exogenously, no a-factor
shown that ste6 mutants produce mature a-factor intracellu- gradient is established (Fig. 4B). Thus, while the MATa cell,
larly (34). We found that exogenous a-factor did not restore which is producing a-factor and therefore establishing a
mating at all when the Aste6 mutant was crossed with pheromone gradient, can effectively court the mfal mfa2
wild-type MATa cells (Table 7). A full interpretation of this mutant, the mfal mfa2 mutant cannot court the MATa cell,
result awaits better characterization of the STE6 gene prod- even though the a-factor is provided exogenously. Instead,
uct, although our results may indicate roles in addition to since there is no a-factor gradient, the MATa cell forms its
a-factor secretion for this protein in the mating process. projection randomly. Our observation that the mating effi-
Perhaps the STE6 gene product is necessary to correctly ciency of wild-type MATa and MATa cells in the presence of
orient internally produced or exogenously supplied a-factor exogenous a-factor is close to that of mfal mfa2 mutants and
in order to facilitate mating. wild-type MATa cells (Table 5) is consistent with the model
An alternative explanation for our observation that exog- shown in Fig. 4. Thus, this model provides a likely explana-
enous a-factor does not restore mating efficiency of mfal tion of why exogenous a-factor does not restore mating
mfa2 mutants to a wild-type level can be made by consider- efficiency of mfal mfa2 mutants to a wild-type level.
ing the results of Jackson and Hartwell (29, 30) on the Our results also show that in the presence of exogenous
VOL. 11, 1991 a-FACTOR STRUCTURE-ACTIVITY RELATIONSHIPS 3611

a-factor, sst2 MATa mutants mate with mfal mfa2 mutants 10. Ciejek, E., and J. Thorner. 1979. Recovery of S. cerevisiae
with about 60-fold greater efficiency than do wild-type a-cells from G1 arrest by alpha-factor pheromone requires
(SS72') MATa cells. However, this result must be inter- endopeptidase action. Cell 18:623-635.
preted cautiously. We were not able to test the effect of 11. Clarke, S. J., J. P. Vogel, R. J. Deschenes, and J. Stock. 1988.
higher doses of a-factor on the mating efficiency of wild-type Posttranslational modification of the Ha-ras oncogene protein:
evidence for a third class of protein carboxyl methyltransferase.
MATa cells to mfal mfa2 mutants because of the extremely Proc. Natl. Acad. Sci. USA 85:4643-4647.
limited solubility of the pheromone. Thus, it is possible that 12. Cross, F., L. H. Hartwell, C. Jackson, and J. B. Konopka. 1988.
higher doses of a-factor might have increased the degree of Conjugation in Saccharomyces cerevisiae. Annu. Rev. Cell
mating restoration of wild-type MATa cells to a level closer Biol. 4:429-457.
to that of sst2 MATa cells. 13. Deschenes, R. J., J. B. Stimmel, S. Clarke, J. Stock, and J. R.
The results reported herein reveal several important rela- Broach. 1989. RAS2 protein of Saccharomyces cerevisiae is
tionships between the structure of a-factor and its biological methyl-esterified at its carboxyl terminus. J. Biol. Chem. 264:
activity and provide evidence that the pheromone works 11865-11873.
exogenously. Recently, studies have appeared which delin- 14. Farnsworth, C. C., M. H. Gelb, and J. A. Glomset. 1990.
Identification of geranylgeranyl-modified proteins in HeLa cells.
eate the steps occurring during the posttranslational proc- Science 247:320-322.
essing of this farnesylated, carboxy-terminal methyl-esteri- 15. Fehrenbacher, G., P. Perry, and J. Thorner. 1978. Cell-cell
fied peptide (28, 39, 41, 55). We are presently investigating in recognition in Saccharomyces cerevisiae: regulation of mating-
detail the specific interaction of the lipopeptide a-factor with specific adhesion. J. Bacteriol. 134:893-901.
its receptor and are searching for antagonists and superac- 16. Finegold, A. A., W. R. Schafer, J. Rine, M. Whiteway, and F.
tive analogs. It is evident that, together with the a-factor, the Tamanoi. 1990. Common modification of trimeric G-proteins
a-factor represents a powerful paradigm for studying cell-cell and ras protein: involvement of polyisoprenylation. Science
communication and for learning about peptide hormone 249:165-169.
action. We anticipate that this system will provide further 17. Fugino, M., C. Kitada, Y. Sakagami, A. Isogai, S. Tamura, and
A. Suzuki. 1980. Biological activity of synthetic analogs of
insights, some possibly unforeseen, into the role of prenyla- tremerogen A-10. Naturwissenschaften 67:406-408.
tion and carboxy-terminal methyl esterification in the bio- 18. Glomset, J. A., M. H. Gelb, and C. C. Farnsworth. 1990. Prenyl
chemistry of eukaryotic organisms. proteins in eukaryotic cells: a new type of membrane anchor.
Trends Biochem. Sci. 15:139-142.
ACKNOWLEDGMENTS 19. Goldstein, J. L., and M. S. Brown. 1990. Regulation of the
mevalonate pathway. Nature (London) 343:425-430.
We thank Wolfgang Duntze, Lorraine Marsh, Susan Michaelis, 20. Goodman, M., J. Coddington, D. F. Mierke, and W. D. Fuller.
and John McGrath for S. cerevisiae strains and Greg Abel for 1987. A model for the sweet taste of sterioisomeric retro-inverso
technical assistance. and dipeptide amides. J. Am. Chem. Soc. 109:4712-4714.
This work was supported by Public Health Service grants GM- 21. Hagen, D. C., and G. F. Sprague. 1984. Induction of the yeast
22086 and GM-22087 from the National Institutes of Health. a-specific STE3 gene by the peptide pheromone a-factor. J. Mol.
Biol. 178:835-852.
REFERENCES 22. Hancock, J. F., A. I. Magee, J. E. Childs, and C. J. Marsh. 1989.
1. Achstetter, T. 1989. Regulation of a-factor production in Sac- All ras proteins are polyisoprenylated but only some are palmi-
charomyces cerevisiae: a-factor pheromone-induced expression toylated. Cell 57:1167-1177.
of the MFal and STE13 genes. Mol. Cell. Biol. 9:4507-4514. 23. HartweUl, L. H. 1973. Synchronization of haploid yeast cell
2. Anderegg, R. J., R. Betz, S. A. Carr, J. W. Crabb, and W. cycles, a prelude to conjugation. Exp. Cell Res. 76:111-117.
Duntze. 1988. Structure of the Saccharomyces cerevisiae mating 24. HartweUl, L. H. 1980. Mutants of Saccharomyces cerevisiae
hormone a-factor: identification of S-farnesyl cysteine as a unresponsive to cell division control by polypeptide mating
structural component. J. Biol. Chem. 263:18236-18240. hormone. J. Cell Biol. 85:811-822.
3. Behling, R. W., and L. W. Jelinski. 1990. Importance of the 25. Herskowitz, I. 1988. Life cycle of the budding yeast Saccharo-
membrane in ligand-receptor interactions. Biochem. Pharma- myces cerevisiae. Microbiol. Rev. 52:536-553.
col. 40:49-54. 26. Herskowitz, I. 1989. A regulatory hierarchy for cell specializa-
4. Brake, A., C. Brenner, R. Najarian, P. Laybourne, and J. tion in yeast. Nature (London) 342:749-757.
Merryweather. 1985. Structure of genes encoding precursors of 27. Horecka, J., and G. F. Sprague (University of Oregon). 1990.
the yeast peptide mating pheromone a-factor, p. 103-108. In Personal communication.
M. J. Gething (ed.), Protein transport and secretion. Cold 28. Hrycyna, C. A., and S. Clarke. 1990. Farnesyl cysteine C-ter-
Spring Harbor Laboratory, Cold Spring Harbor, N.Y. minal methyltransferase activity is dependent upon the STEJ4
5. Bucking-Throm, E., W. Duntze, L. H. Hartwell, and T. R. gene product in Saccharomyces cerevisiae. Mol. Cell. Biol.
Manney. 1973. Reversible arrest of haploid yeast cells at the 10:5071-5076.
initiation of DNA synthesis by a diffusible sex factor. Exp. Cell 29. Jackson, C. L., and L. H. Hartwell. 1990. Courtship in Saccha-
Res. 76:99-110. romyces cerevisiae: an early cell-cell interaction during mating.
6. Casey, P. J., P. A. Solski, C. J. Der, and J. E. Buss. 1989. p2lras Mol. Cell. Biol. 10:2202-2213.
is modified by a farnesyl isoprenoid. Proc. Natl. Acad. Sci. 30. Jackson, C. L., and L. H. Hartwell. 1990. Courtship in Saccha-
USA 86:8323-8327. romyces cerevisiae: both cell types choose mating partners by
7. Chan, R. K., and C. A. Otte. 1982. Isolation and genetic analysis responding to the strongest pheromone signal. Cell 63:1039-
of Saccharomyces cerevisiae mutants supersensitive to G, 1051.
arrest by a factor and a factor pheromones. Mol. Cell. Biol. 31. Kamiya, Y., A. Sakurai, S. Tamura, and N. Takahashi. 1978.
2:11-20. Structure of rhodotorucine A, a novel lipopeptide, inducing
8. Chan, R. K., and C. A. Otte. 1982. Physiological characteriza- mating tube formation in Rhodosporidium toruloides. Biochem.
tion of Saccharomyces cerevisiae mutants supersensitive to G, Biophys. Res. Commun. 83:1077-1083.
arrest by a-factor and a-factor pheromones. Mol. Cell. Biol. 32. King, K., H. G. Dohlman, J. Thorner, M. G. Caron, and R. J.
2:21-29. Lefkowitz. 1990. Control of yeast mating signal transduction by
9. Chang, F., and I. Herskowitz. 1990. Identification of a gene a mammalian P2-adrenergic receptor and G, a subunit. Science
necessary for cell cycle arrest by a negative growth factor of 250:121-123.
yeast: FAR1 is an inhibitor of a G1 cyclin, CLN2. Cell 63:999- 33. Kronstad, J. W., J. A. Holly, and V. L. MacKay. 1987. A yeast
1011. operator overlaps an upstream activation site. Cell 50:369-377.
3612 MARCUS ET AL. MOL. CELL. BIOL.

34. Kuchler, K., R. E. Sterne, and J. Thorner. 1989. Saccharomyces New York.
cerevisiae STE6 gene product: a novel pathway for protein 50. Rilling, H. C., E. Breunger, W. W. Epstein, and P. F. Crain.
export in eukaryotic cells. EMBO J. 8:3973-3984. 1990. Prenylated proteins: the structure of the isoprenoid group.
35. Kurjan, J. 1985. a-Factor structural gene mutations in Saccha- Science 247:318-320.
romyces cerevisiae: effects on a-factor production and mating. 51. Sakagami, Y., A. Isogal, A. Suzuki, S. Tamura, C. Kitada, and
Mol. Cell. Biol. 5:787-796. M. Fujino. 1979. Structure of tremerogen A-10, a peptidyl
36. Lipke, P. N., A. Taylor, and C. E. Ballou. 1976. Morphogenetic hormone inducing conjugation tube formation in Tremella me-
effect of a-factor on Saccharomyces cerevisiae. J. Bacteriol. senterica. Agric. Biol. Chem. 43:2643-2645.
127:610-618. 52. Sakagami, Y., M. Yoshida, A. Isogal, and A. Suzuki. 1981.
37. Maltese, W. A., and R. A. Erdman. 1989. Characterization of Structure of tremerogen a-13, a peptidyl hormone of Tremella
isoprenoid involved in the post-translational modification of mesenterica. Agric. Biol. Chem. 45:1045-1047.
mammalian cell proteins. J. Biol. Chem. 264:18168-18172. 53. Sargent, D. F., and R. Schwyzer. 1986. Membrane lipid phase as
38. Maltese, W. A., and J. D. Robishaw. 1990. Isoprenylation of catalyst for peptide-receptor interactions. Proc. Natl. Acad.
C-terminal cysteine in a G-protein -y subunit. J. Biol. Chem. Sci. USA 83:5774-5778.
265:18071-18074. 54. Schafer, W. R., R. Kim, R. Sterne, J. Thorner, S.-H. Kim, and
39. Marcus, S., G. A. Caldwell, C.-B. Xue, F. Naider, and J. M. J. Rine. 1989. Genetic and pharmacological suppressors of
Becker. 1990. Total in vitro maturation of the Saccharomyces oncogenic mutations in RAS genes of yeast and humans. Sci-
cerevisiae a-factor lipopeptide mating pheromone. Biochem. ence 245:379-385.
Biophys. Res. Commun. 172:1310-1316. 55. Schafer, W. R., C. E. Trueblood, C.-C. Yang, M. P. Mayer, S.
40. Marcus, S., C.-B. Xue, F. Naider, and J. M. Becker. 1991. Rosenberg, C. D. Poulter, S.-H. Kim, and J. Rine. 1990. Enzy-
Degradation of a-factor by a Saccharomyces cerevisiae a mat- matic coupling of cholesterol intermediates to a mating phero-
ing type-specific endopeptidase: evidence for a role in recovery mone precursor and to the Ras protein. Science 249:1133-1139.
of cells from G1 arrest. Mol. Cell. Biol. 11:1030-1039. 56. Steden, M., R. Betz, and W. Duntze. 1989. Isolation and
41. Marr, R. S., L. C. Blair, and J. Thorner. 1990. Saccharomyces characterization of Saccharomyces cerevisiae mutants super-
cerevisiae STEJ4 gene is required for COOH-terminal methyl- sensitive to GI arrest by the mating hormone a-factor. Mol.
ation of a-factor mating pheromone. J. Biol. Chem. 265:20057- Gen. Genet. 219:439-444.
20060. 57. Tkacz, J. S., and V. L. MacKay. 1979. Sexual conjugation in
42. McCaffrey, G., F. J. Clay, K. Kelsay, and G. F. Sprague, Jr. yeast. Cell surface changes in response to the action of mating
1987. Identification and regulation of a gene required for cell pheromones. J. Cell Biol. 80:326-333.
fusion during mating of the yeast Saccharomyces cerevisiae. 58. Trueheart, J., J. D. Boeke, and J. R. Fink. 1987. Two genes
Mol. Cell. Biol. 7:2680-2690. required for cell fusion during yeast conjugation: evidence for a
43. McGrath, J. P., and A. Varshavsky. 1989. The yeast STE6 gene pheromone-induced surface protein. Mol. Cell. Biol. 7:2316-
encodes a homologue of the mammalian multidrug resistance 2328.
P-glycoprotein. Nature (London) 340:400-404. 59. Tsuchiya, E., S. Fukui, Y. Kamiya, Y. Sakagami, and M. Fugino.
44. Michaelis, S., and I. Herskowitz. 1988. The a-factor pheromone 1978. Requirement of chemical structure of lipopeptidyl factors
of Saccharomyces cerevisiae is essential for mating. Mol. Cell. inducing sexual differentiation on vegetative cells of heteroba-
Biol. 8:1309-1318. sidiomycetous yeasts. Biochem. Biophys. Res. Commun. 85:
45. Miller, J. H. 1972. Experiments in molecular genetics. Cold 459-463.
Spring Harbor Laboratory, Cold Spring Harbor, N.Y. 60. Van Ardsdell, S. W., G. L. Stetler, and J. Thorner. 1987. The
46. Miyakawa, T., M. Nishihara, E. Tsuchiya, and S. Fukui. 1982. yeast repeated element sigma contains a hormone-inducible
Role of metabolism of the mating pheromone in the sexual promoter. Mol. Cell. Biol. 7:749-759.
differentiation of the heterobasidiomycete Rhodosporidium 61. Vorburger, K., G. T. Kitten, and E. A. Nigg. 1989. Modification
toruloides. J. Bacteriol. 151:1184-1194. of nuclear lamin proteins by a mevalonic acid derivative occurs
47. Naider, F., and J. M. Becker. 1986. Structure-activity relation- in reticulocyte lysates and requires the cysteine residue of the
ships of the yeast alpha-factor. Crit. Rev. Biochem. 21:225-248. C-terminal CXXM motif. EMBO J. 8:4007-4013.
48. Powers, S., S. Michaelis, D. Broek, S. S. Anna-A., J. Field, I. 62. Xue, C.-B., G. A. Caldwell, J. M. Becker, and F. Naider. 1989.
Herskowitz, and M. Wigler. 1986. RAM, a gene of yeast required Total synthesis of the lipopeptide a-mating factor of Saccharo-
for a functional modification of RAS proteins and for production myces cerevisiae. Biochem. Biophys. Res. Commun. 162:253-
of mating pheromone a-factor. Cell 47:413-422. 257.
49. Reynolds, A., and V. Lundblad. 1989. Assay for 3-galactosidase 63. Xue, C.-B., A. Ewenson, J. M. Becker, and F. Naider. 1990.
in liquid cultures, p. 13.6.1-13.6.4. In Ausubel et al. (ed.), Solution phase synthesis of Saccharomyces cerevisiae a-mating
Current protocols in molecular biology. Wiley Interscience, factor and its analogs. Int. J. Peptide Protein Res. 36:362-373.