Previous Article | Next Article 
Mol Cell Biol, January 1998, p. 221-232, Vol. 18, No. 1
0270-7306/98/$04.00+0
Copyright © 1998, American Society for Microbiology. All rights reserved.
Dual Sets of Chimeric Alleles Identify Specificity
Sequences for the bE and bW Mating and
Pathogenicity Genes of Ustilago maydis
A. R.
Yee1 and
J. W.
Kronstad1,2,*
Departments of Microbiology and
Immunology2 and
Plant
Science,1 Biotechnology
Laboratory, University of British Columbia, Vancouver, British
Columbia V6T 1Z3, Canada
Received 10 July 1997/Returned for modification 9 September
1997/Accepted 28 October 1997
 |
ABSTRACT |
The b mating-type locus of the fungal plant pathogen
Ustilago maydis encodes two multiallelic gene products, bE
and bW, that control the formation and maintenance of the infectious
cell type. Dimerization via the N-terminal regions of bE and bW
proteins encoded by alleles of different specificities establishes a
homeodomain-containing transcription factor. The bE and bW products
encoded by alleles of like specificities fail to dimerize. We
constructed sets of chimeric alleles for the bE1 and
bE2 genes and for the bW1 and bW2
genes to identify sequences that control specificity. The mating
behavior of strains carrying chimeric alleles identified three classes
of specificity: b2 (class I), specificity different from
either parental type (class II), and b1 (class III).
Crosses between strains carrying bE and bW
chimeric alleles identified two short blocks of amino acids that
influence specificity and that are located in the N-terminal variable
regions of the b proteins. Comparisons of pairs of chimeric alleles
encoding polypeptides differing in specificity and differing at single
amino acid positions identified 16 codon positions that influence the
interaction between bE and bW. Fifteen of these positions lie within
the blocks of amino acids identified by crosses between the strains
carrying chimeric alleles. Overall, this work provides insight into the organization of the regions that control recognition.
| |
INTRODUCTION |
Recognition mediated by
protein-protein interactions plays a fundamental role in many
biological processes. Well-characterized examples include
antibody-antigen interactions (8, 9, 23), ligand-receptor
binding (22, 35), and the establishment and maintenance of
tissue integrity by cadherins (19). The proteins involved in
sexual reproduction and incompatibility in fungi provide relatively
simple examples of determinants of self versus nonself recognition. In
this paper, we describe a molecular genetic approach to identify the
determinants of recognition for the proteins encoded by the
b mating-type locus of the fungal corn pathogen
Ustilago maydis.
U. maydis is commonly found in nature as black diploid
teliospores on infected corn plants (6). The teliospores
germinate, and meiosis occurs to produce haploid, yeast-like progeny.
Nonself recognition between compatible haploid mating partners is a
prerequisite to the establishment of an infectious, dikaryotic cell
type, and the genes at the a and b mating-type
loci are considered pathogenicity factors (reviewed in references
2 and 18). The a
locus, with alternate specificities a1 and a2,
encodes pheromones and pheromone receptors and controls recognition of
mating partners at the level of cell fusion (3, 11, 31). The
b locus controls the formation and maintenance of the
infectious cell type after cell fusion has occurred. If the cells
participating in mating have different specificities (nonself) at the
b locus, a vigorous, straight dikaryotic filament is formed
and this cell type will be infectious. In contrast, mating partners
that carry b sequences of like specificities (self) do not
form an infectious dikaryon. Interestingly, the b locus is
believed to have at least 25 different naturally occurring specificities (24, 29), and all of the nonself combinations of alleles are able to promote pathogenicity and sexual development.
The b locus of U. maydis was initially cloned by
transformation of a library of DNA from a strain with b1
specificity into a diploid strain with b2 specificity and by
subsequent screening of transformants for filamentous growth
(16). The molecular characterization of the b
locus revealed the presence of two divergently transcribed genes called
bE (encoding a polypeptide of 473 amino acids) and
bW (encoding a polypeptide of 644 amino acids) (12, 17,
28). These genes exist in an allelic series such that each of the
25 specificities at the b locus is determined by the specific bE and bW alleles present in a haploid
strain. The bE and bW gene products do not show
sequence similarity to each other except that each contains a
homeodomain-like region of approximately 60 amino acids that lies
between a variable amino-terminal region (N-terminal region; 100 to 150 amino acids) and a conserved carboxy-terminal region (C-terminal
region) (12, 28). Gene disruption experiments revealed that
the b gene products are necessary to establish the filamentous dikaryon. That is, strains compatible at the a
locus but carrying null mutations in both bE and
bW are defective in mating (12, 17). Furthermore,
deletion of the b genes revealed that the presence of one
bE and one bW from each mating partner (e.g.,
bE1 plus bW2 or bE2 plus
bW1) is sufficient to allow mating and pathogenic
development in the plant (12).
It is believed that any combination of bE and bW
gene products encoded by different alleles is capable of triggering
dikaryon formation. In contrast, the bE and bW
products from the same strain fail to initiate pathogenic development.
Experiments using the two-hybrid system with Saccharomyces
cerevisiae and an in vitro protein binding assay indicate that the
N-terminal regions of bE and bW promote dimerization between gene
products from alleles of different specificities (15). The
bE and bW products from genes of the same strain
(e.g., bE2 and bW2) fail to dimerize. Thus, it
appears that the variable regions of bE and bW are dimerization domains
and that the formation of active heterodimers requires b polypeptides
from genes with different specificities. In this context, the critical
question is the following: what mechanism prevents the dimerization of
bE and bW gene products from the same locus?
In previous work, we defined a specificity region for the
bE1 and bE2 genes by the construction and
analysis of chimeric alleles (37). This work identified a
40-amino-acid sequence within the N-terminal variable region that was
thought to contain the residues that control specificity. Surprisingly,
chimeric alleles between bE1 and bE2 that
contained recombination sites within the sequence encoding the
40-amino-acid region displayed specificities different from that of
either parental bE allele. These alleles were designated class II to distinguish them from alleles that had not changed specificity (class I) or that had switched specificity from one parental type to the other (class III). Kämper et al. have shown that single-amino-acid changes within a similar portion of the variable
region of bE2 allow dimerization with bW2 (15). These observations prompted the proposal that a dimerization interface mediates the attraction between bE and bW proteins from different alleles. In this context, bE and bW proteins that fail to dimerize must
fail to do so because of key interfering residues that block association by preventing productive interactions or by establishing disruptive interactions, e.g., by polar or hydrophobic effects or by
steric hindrance (14).
In this paper, we report the construction of additional chimeric
alleles for the bE1 and bE2 genes and the
construction of a large set of chimeric alleles for the bW1
and bW2 genes. Overall, these sets of alleles provided a
refined view of the 40-amino-acid specificity region for bE1
and bE2 and identified an analogous 70-amino-acid region for
the bW1 and bW2 alleles. In addition, crosses
between all combinations of bE and bW chimeric
alleles revealed that the borders of the regions defined by class II
alleles contain the important determinants for recognition. For
bE1 and bE2 alleles, the specificity borders lie
between codons 31 and 39 and between codons 79 and 92; for the
bW1 and bW2 alleles, these borders are found
between codons 2 and 9 and 74 and 83. Our data suggest that the
40-amino-acid region for bE and the 70-amino-acid region for
bW represent the intervals between the specificity
determinants in the border regions. Key amino acid positions within the
borders were identified by comparisons of chimeric alleles that
differed at a single codon and had different specificities when tested
against strains carrying either wild-type or chimeric alleles.
Additional chimeric alleles, constructed between bW1 and
bW3, indicated that a single border region can be sufficient
to control the interaction for certain allele pairs.
(This work fulfills part of the requirements for a Ph.D. in plant
science from the University of British Columbia for A. R. Yee.)
| |
MATERIALS AND METHODS |
Strains and growth conditions.
Escherichia coli DH5
[F
endA1 hsdR17 (rK
mK+) supE44 Thi
1
recA1 
80dlac ZM15] (Bethesda Research
Laboratories) was used for DNA manipulations and was grown in
Luria-Bertani medium (26). U. maydis wild-type
prototrophic strains were 518 (a2 b2), 521 (a1
b1), 031 (a1 b2), and 032 (a2 b1)
(16). The genotypes of strains carrying chimeric
b alleles (for bE1 and bE2 or
bW1 and bW2) were designated bEx and
bWx where x is followed by a number indicating
the amino acid position corresponding to the codon at which the
sequence changes from E1 to E2 or W1
to W2. The genotypes for chimeric b alleles for
bW1 and bW3 were distinguished from those for the
bW1 and bW2 alleles by the designation
bW1/3x followed by the codon number at which the sequence
changes from bW1 to bW3. It should be noted that
the procedure for constructing chimeric alleles disrupts the wild-type
allele of bE or bW in each strain due to the
insertion of a hygromycin resistance cassette and leaves only one
functional b gene (the chimeric gene) (37).
U. maydis cultures were grown in potato dextrose medium
(Difco Laboratories) or complete medium (13), and mating
reactions were carried out on solid double-complete medium containing
1% activated charcoal (13).
Construction of chimeric alleles.
The strategy for the in
vivo construction of chimeric alleles by transformation and homologous
integration was described previously (37). This approach
involved the transformation of deletion derivatives of bE
and bW genes into strains of opposite b
specificities and the generation of chimeric alleles by homologous
integration at b. Additional chimeric alleles were
constructed in vitro by a PCR approach and subsequently used to replace
wild-type alleles by transformation. The PCR approach was based on the
Megaprimer PCR procedure (27) and is diagrammed in Fig.
1. This technique requires two rounds of
PCR amplification and a "chimeric" primer designed to overlap the
junction between the bE1 and bE2 or
bW1 and bW2 sequences. The first round involves
PCR with the chimeric primer and a second primer to produce a PCR
product called the "megaprimer." This round of PCR employed
bE2 or bW2 sequences as templates. The megaprimer
was then used in a second round of PCR with a third primer and
bE1 or bW1 template DNA to produce the final
chimeric product. The primers employed for constructing chimeric
alleles are shown in Table 1, and the
chimeric alleles are listed in Table 2.
View larger version (20K):
[in this window]
[in a new window]
|
FIG. 1.
In vitro construction of chimeric alleles. The
megaprimer PCR method (27) was employed to construct
chimeric alleles of bE1 and bE2 and
bW1 and bW2. The procedure is diagrammed for the
construction of bE1 and bE2 chimeric alleles; the
same approach was used for bW1 and bW2 chimeric
alleles except that the final chimeric amplification product was cloned
into plasmid pAR69 (see Materials and Methods). The chimeric primer is
designed to overlap the anticipated junction point of the chimeric
allele. The megaprimer approach makes use of two additional, flanking
primers. For bE chimeric alleles, primers BE10 and bW1-Y6A
were used to produce the megaprimer and to isolate a fragment
containing 5' promoter sequences, respectively. For bW
chimeric alleles, the equivalent primers were BW6 and BE1-Y31A
(Materials and Methods). Note that the final transformation cassette
disrupts bE or bW upon homologous integration and
leaves the chimeric allele as the only functional b gene in
the transformant.
|
|
The protocol of Sarkar and Sommer (
27) was employed for PCR
except that Vent polymerase (New England BioLabs, Beverly, Mass.)
was
used instead of
Taq polymerase. The reaction mixture (100
µl) for PCR with Vent polymerase, as recommended by Cease et al.
(
5), contained 10 mM KCl, 10 mM
(NH
4)
2SO
4, 20 mM Tris-HCl (pH
8.8 at 25°C), 3 mM MgSO
4, a 500 µM concentration of each
dNTP,
0.1% Triton X-100, 1 µM concentrations of primers, 1 fmol of
DNA
template, and 1 U of Vent polymerase. Thermal cycling was done
with
a Perkin-Elmer 480 with an initial 3-min time delay at 94°C
and a
step cycle of 1 min at 94°C, 1 min at 52°C, and 1 min at
72°C;
the samples were held for 10 min at 72°C at the end of the
cycles.
The chimeric PCR products were first cloned as blunt-end fragments into
pBluescript KS+ (Stratagene) which had been linearized
with
EcoRV. The cloned chimeric or mutant product was then
subcloned
from pBluescript KS+ into a
Ustilago
transformation construct.
For
bE, this transformation
construct was pMBE2 (Fig.
1) which
had been linearized with
SacI, made blunt with T4 polymerase,
and dephosphorylated.
Plasmid pMBE2 contains a 1.8-kb
BglII-
SalI
fragment carrying the hygromycin resistance cassette in pUC9
(
34).
The plasmid also contains a 1.4-kb fragment encoding
the N-terminal
portion of the bW1 polypeptide. For
bW
chimeras, the transformation
construct was pAR69 which had been
linearized with
HindIII, made
blunt with T4 polymerase, and
dephosphorylated. Plasmid pAR69
is based on pBluescript KS+ and
contains a 1.9-kb
BglII-
XbaI fragment
encoding
the hygromycin resistance cassette (
34) and a 5.8-kb
XbaI-
BamHI fragment encoding the
bE1
gene and 3' flanking sequences.
Subclones containing the correct
orientation of insert were chosen
for transformation into
U. maydis. The
bE and
bW chimeric allele
constructs were linearized with
BamHI, extracted once with
phenol-chloroform-isoamyl
alcohol (24:24:1) and once with
chloroform-isoamyl alcohol (24:1),
ethanol precipitated, and dissolved
in Tris-EDTA. This DNA was
then used to transform
U. maydis
protoplasts. Transformation of
U. maydis was accomplished by
a protoplast-polyethylene glycol-CaCl
2 procedure modified
from Wang et al. (
34) and Specht et al. (
30).
Homologous integration and replacement of sequences at the
b
locus
were confirmed by DNA hybridization (data not shown).
Mating tests.
Routine mating tests employed a
"drop-on-drop" procedure. Overnight cultures of U. maydis cells were grown at 30°C and 225 rpm for 16 to 20 h
until late log or early stationary phase (optical density at 600 nm,
1.8 to 2.2). Then, 10 to 30 µl of culture was dropped on a charcoal
mating plate containing 1% glucose and allowed to dry. A second drop
of the tester culture was placed on top of the first and allowed to
dry. The plates were then taped with a double layer of Parafilm,
incubated in the dark at room temperature for 24 to 48 h, and
scored for mycelial growth.
| |
RESULTS |
Chimeric alleles of bE1 and bE2.
Previously,
we reported the construction of 16 chimeric alleles of the
bE1 and bE2 genes (37). The strategy
for constructing chimeric alleles at the b locus involves
transformation of U. maydis cells with truncated
bE or bW genes such that chimeric alleles are
generated as a result of homologous recombination within the variable
5' portion of the gene (37). In addition, it was also
possible to construct chimeric alleles in vitro and to replace
wild-type alleles by transformation and homologous recombination. In
each case, the chimeric alleles were constructed such that deletions or
insertions did not occur at the point of recombination between
sequences from different alleles, as confirmed by sequence analysis.
Our earlier work on chimeric
bE alleles identified three
classes based on the mating activity of the host strains; those that
had a
bE2 specificity (class I), those with a
bE1
specificity
(class III) and those with a specificity different from
bE1 or
bE2 (class II). The determination of the
positions of the recombination
sites for these classes identified a
region involved in specificity
between codons 39 and 87 which encodes a
portion of the N-terminal
variable region. We have now obtained five
new chimeric alleles
for
bE1 and
bE2
(
bEx31,
bEx45,
bEx57,
bEx82,
bEx89) to develop
a more detailed map of
the three specificity classes for the chimeric
bE alleles.
The new chimeric alleles are shown with the previously
constructed
alleles on a map of the
bE1 and
bE2 variable
region
in Fig.
2A. It should be noted
that allele
bEx39 was previously
scored as having a
specificity like that of
bE2 (
37); the subsequent
isolation of additional strains carrying this allele and further
incompatibility tests revealed mating activity with both
bW1
and
bW2 tester strains. With the additional chimeric alleles
for
bE1 and
bE2, we have now obtained chimeras
for 17 of the 36 potential
positions in the variable region between
codons 1 and 107. These
chimeras include 15 of the 27 potential
chimeras in the first
92 codons that encode the N- and C-terminal
borders of the region
identified by the analysis of the class II
alleles.
View larger version (19K):
[in this window]
[in a new window]
|
FIG. 2.
Map of the specificity regions for chimeric alleles of
bE1 and bE2 and chimeric alleles of
bW1 and bW2. (A) The positions of recombination
sites for the sequence encoding the N-terminal variable region
(positions 1 to 160 of the 473-amino-acid polypeptide) are shown for 21 chimeric alleles. The numbers on the map indicate the codons at which
each recombination event changed the coding sequence from
bE1 (N terminal) to bE2 (C terminal). The mating
reactions of strains carrying the chimeric alleles are indicated below
the map. The strains carrying chimeric alleles have a2
specificity and were tested against strains with a1 b1 and
a1 b2 specificity. A plus sign indicates a compatible mating
reaction that results in the formation of white aerial hyphae on mating
colonies. A minus sign indicates a failure of the mating mixture to
form aerial hyphae. Three classes of mating behavior are exhibited by
the strains carrying chimeric alleles: class I (bE2
specificity), class II (novel specificity), and class III
(bE1 specificity). The newly constructed chimeric alleles
are marked with an asterisk to distinguish them from the alleles
described previously (37). (B) The positions of
recombination sites for the sequence encoding the N-terminal variable
region (positions 1 to 120 of the 644-amino-acid polypeptide) are shown
for 24 chimeric alleles. The numbers on the map indicate the codons at
which each recombination event changed the coding sequence from
bW1 (N terminal) to bW2 (C terminal). The mating
reactions of strains carrying the chimeric alleles are indicated below
the map. The strains carrying chimeric alleles have a1
specificity and were tested against strains with a2 b1 and
a2 b2 specificities (see Fig. 4). As with the strains
carrying the bE chimeric alleles shown in panel A, three
classes of mating behavior were found: class I (bW2
specificity), class II (novel specificity), and class III
(bW1 specificity).
|
|
Chimeric alleles of bW1 and bW2.
Gillissen
et al. (12) presented genetic evidence that the specificity
of recognition determined by b is mediated by interactions between the bE and bW gene products from strains
with different b specificities rather than by bE-bE or bW-bW
interactions. In addition, our previous analysis of bE
chimeric alleles indicated that each allele with novel specificity
(class II alleles) was found to have identical mating behavior when
tested against a set of strains carrying naturally occurring
bW alleles (37). Therefore, we constructed a set
of chimeric alleles for bW1 and bW2 to further
explore the interaction between bE and bW and to attempt to collect bW chimeric alleles that might identify
differences between class II bE chimeric alleles. As shown
in Fig. 2B, chimeric alleles were obtained for 24 of the 43 potential
positions (56%) between codons 1 and 109 of bW1 and
bW2. The potential positions for the formation of chimeric
alleles represent the codons that specify different amino acids in bW1
and bW2. Mating tests with strains carrying wild-type bE1
and bE2 alleles revealed that the transformants carrying the
chimeric bW alleles represented three classes:
bW2 (class I), specificity different from bW1 and
bW2 (class II), and bW1 (class III) (Fig.
3). A sequence analysis of the
bW chimeric alleles from each class allowed the
identification of a region involved in determining specificity
comparable to that found for bE1 and bE2 and
located between codons 6 and 83 (the N-terminal variable region for bW
is encoded by codons 1 to 150). Interestingly, the region between
codons 76 and 83 did not show a distinct transition between chimeric
alleles that had a novel specificity (class II) and those that had a
bW1 (class III) specificity. This feature suggests that
amino acids involved in determining specificity may be clustered in the
part of the variable region specified by these codons. The C-terminal
border of the region defined by the class II alleles of bE
did not show a similar complexity (Fig. 2A) (37).
View larger version (20K):
[in this window]
[in a new window]
|
FIG. 3.
Mating reactions between strains carrying wild-type
alleles and bW chimeric alleles. Representative mating
reactions between strains with chimeric alleles and strains with
wild-type alleles (032 [a2 b1] and 518 [a2
b2]) to demonstrate the activity of the chimeric alleles shown in
Fig. 2. Each colony develops from the coinoculation of the two strains
to be tested on medium containing activated charcoal (to enhance the
reaction). Positive controls (white aerial hyphae) for mating are shown
for the interactions of wild-type strains 032 (a2 b1) and
031 (a1 b2) and 521 (a1 b1) and 518 (a2
b2). Negative controls (flat gray colonies) include the
interaction of 031 (a1 b2) with 518 (a2 b2) and
032 (a2 b1) with 521 (a1 b1); these strains,
although capable of fusion, have the same specificity at b.
Note that certain allele combinations give weak mating reactions:
bWx79 and bWx82 with bE1 (032).
|
|
During the construction of the
bW chimeric alleles, a
specific attempt was made to obtain a large number of recombinant
alleles
in the borders of the region defined by the class II alleles.
As a result, chimeric alleles were obtained for 10 of the 12 potential
positions (83%) between codons 1 and 52 (encoding the N-terminal
border region) and for 14 of the 18 potential positions (77%)
between
codons 68 and 109 (encoding the C-terminal border region).
As described
above, the potential positions for generating chimeric
alleles are the
codons for bW1 that differ from those for bW2.
These chimeric alleles
were initially obtained to provide a detailed
analysis of the regions
of transition between classes with different
specificities. As
described below, however, these chimeric alleles
have also allowed the
identification of specific amino acid positions
that control
recognition between
bE and
bW gene products.
Chimeric alleles of bW1 and bW3.
To date,
our chimeric allele analysis has focused on bE and
bW genes of b1 and b2 specificities.
We were interested in expanding the analysis to include additional
b specificities to determine whether similar sequences in
the N-terminal variable regions were important in different allele
combinations. The b3 locus provided a straightforward
starting point for this analysis because alignments of the predicted
amino acid sequences encoded by bW1 and bW3
revealed that the variability between these alleles shows up primarily in the first 20 amino acids at the corresponding N termini (Fig. 4A). Therefore, this region potentially
contains all of the specificity determinants for this allele pair.
Interestingly, differences exist at only 5 of the first 20 positions
and at only 4 of the remaining 140 amino acids in the N-terminal region
upstream of the homeodomain. The sequence similarity suggests that
bW1 and bW3 are evolutionarily close, relative to
other allele pairs, and that only a few differences in one region are
sufficient to change specificity.
View larger version (16K):
[in this window]
[in a new window]
|
FIG. 4.
Construction of chimeric alleles for bW1 and bW3. (A) An
alignment of the sequences at the N-terminal regions of bW1 and bW3
shows a high degree of identity; amino acid differences are found at
nine positions in the first 160 positions. Alignments of additional bW
sequences have been described previously (12). (B) Three
chimeric alleles (bW1/3x9, -20, and -34) were constructed between
bW1 and bW3; the number following the x indicates
the first position of the bW3 sequence. The amino acid sequences
encoded by the first 40 codons of the chimeric alleles are shown
aligned with the sequences of the comparable regions from bW1 and bW3.
For the chimeric alleles, the sequences before the asterisk are from
the bW1 protein and all three have bW1 specificity.
|
|
Three chimeric alleles were obtained between
bW1 and
bW3 to confirm the position of the specificity determinants
for this
allele pair (Fig.
4B). Each of the chimeric alleles
(
bW1/3x9,
bW1/3x20, and
bW1/3x34) had
class III (
bW1) specificity, and alleles
with class I and II
specificities were not found. The relatively
small region of variable
sequence between
bW1 and
bW3 probably
accounts
for the absence of chimeric alleles of the class I and
II types. The
region for recombination to generate these alleles
would be relatively
small, and it would be necessary to construct
this type of allele in
vitro. The class III specificity of the
alleles between
bW1
and
bW3 indicates that the specificity region
for this
allele pair lies upstream of codon 9, as predicted by
sequence
inspection. Overall, these results indicate that a quite
different map
of the bW N-terminal variable region can be obtained
depending on the
bW and
bE alleles under consideration. However,
in terms of specificity determinants, the
bW1/bW3
combination
revealed that a short N-terminal region is important; this
region
may play an identical role to that of the N-terminal region
identified
for the
bW1/bW2 combination. In particular, codon
6, which plays
an important role in specificity (described below),
encodes an
amino acid in this region and was found to encode different
amino
acids when the bW1 and bW3 polypeptide sequences were compared.
Crosses between strains carrying bE and bW
chimeric alleles.
The availability of sets of bE1/bE2
and bW1/bW2 chimeric alleles presented an opportunity to
further investigate the roles of the specificity regions through
incompatibility tests between strains carrying different chimeric
bE and bW genes. As described above, these sets
of chimeras each defined three specificity classes: class I, wild-type
b2 specificity; class II, novel specificity (different from
b1 or b2); and class III, wild-type b1
specificity. Mating tests on culture medium were performed for all
combinations between strains carrying each of 19 different
bE1/bE2 chimeras and strains carrying each of the 26 different bW1/bW2 chimeras (494 combinations). The results
of these crosses are summarized in Table
2, and representative mating tests are
shown in Fig. 5. These tests provided an
interesting insight into the organization of the specificity regions of
the bE and bW proteins. Specifically, the striking general result was
that the majority of strains carrying chimeric alleles from class II
failed to give a positive mating reaction, suggesting that the borders
of the regions defined by these alleles contain the important residues
for recognition and dimerization between bE and bW. As diagrammed in
Fig. 6, these border sequences have been
designated N1 and C1 for the b1 specificity genes and N2 and
C2 for the b2 specificity genes. The bE and
bW chimeric alleles of class II specificity would therefore
be associated with the N1 and C2 border combination. Note that our
strategy would not yield chimeric alleles encoding the N2 and C1 border combination.
View larger version (108K):
[in this window]
[in a new window]
|
FIG. 5.
Mating reactions between strains carrying wild-type or
chimeric alleles of bE1 and bE2 or bW1
and bW2. Representative mating reactions are shown to
illustrate the data summarized in Table 2. The appearance of vigorous
white aerial hyphae indicates a strongly compatible interaction between
bE and bW polypeptides; this type of reaction (e.g., bW2 with bE1) is
assigned three pluses in Table 2; weaker compatible reactions are
assigned one or two pluses (e.g., bEx90 with bWx76).
|
|
View larger version (13K):
[in this window]
[in a new window]
|
FIG. 6.
Model for the interactions of the borders of the
specificity regions of bE and bW. The borders are designated N and C
followed by a number (1 or 2) to indicate the specificity of the
parental allele. Compatible interactions (dimerization) would align N
and C borders of different specificities (top), and incompatibility
would result from the interaction of borders of like specificity
(second from top). An interaction leading to dimerization is indicated
by the black boxes between borders of different specificities. Chimeric
alleles with recombination points between the N and C borders (e.g.,
class II alleles such as bEx57 and bWx52) give an
incompatible reaction (Table 2) because borders of like specificity are
aligned. In contrast, chimeric alleles of class II specificity (e.g.,
bEx57) give compatible reactions with wild-type alleles
(e.g., bW1). In this case, it is sufficient for either the N
or the C border to be recognized as non-self. A similar situation can
be found for some naturally occurring alleles such as the combination
of bW1 and bW3 where the determinants of
specificity are found only in the N-terminal border region.
|
|
As shown in Fig.
6, the recombination events between the border regions
that generate class II alleles would result in
b gene
products that are capable of interacting with products from either
parental allele. For example, the product of class II allele
bEx57 would allow a positive mating interaction with strains
carrying
bW1 and
bW2 alleles. This suggests that
dimerization is not prevented
when sequences of like specificities are
present at just one of
the borders (e.g., a bE N1/C2 interaction with a
bW N1/C1). Conversely,
recognition of nonself at one border is
sufficient to allow dimerization.
We propose that the products of the
class II alleles fail to interact
with each other because self
combinations are present for both
of the borders (e.g., N1/C2 with
N1/C2). An example of this situation
is depicted in Fig.
6 for class II
alleles
bWx52 and
bEx57. These
combinations would
be similar, in terms of border combinations,
to the wild-type self
combinations of N1/C1 with N1/C1, or N2/C2
with N2/C2. Overall, the
mating tests between strains with chimeric
alleles indicated that the
borders of the specificity intervals
contain amino acid residues that
are important for recognition;
this is consistent with the border
locations of residues that
influence specificity (see below) and the
location of a single
short N-terminal border for the
bW1/bW3
combination described
earlier.
The crosses presented in Table
2 also revealed that the
bEx107,
bEx128, and
bEx156 alleles
were found to apparently have
specificities different from that of the
wild-type
bE1 allele
when tested against
bW
chimeras
bWx76,
bWx77,
bWx80, and
bWx81.
Surprisingly, the
bEx156 allele contains
all of the variable region
of the
bE1 gene (encoding amino
acids 1 to 156) fused to a portion
of the
bE2 gene encoding
part of the C-terminal region (amino
acids 157 to 473). A comparison of
the predicted amino acid sequences
of the products of the
bEx156 and
bE1 alleles revealed three differences
in the homeodomain and one difference in the C-terminal region.
It is
possible that these residues in the homeodomain contribute
to the
specificities of interactions in other allele combinations
because
variability was found in this region when the sequences
of six
bE alleles were compared (
17). This result
suggests a
possible role for the homeodomain in specificity that is
only
revealed through test crosses with specific
bW chimeric
alleles.
It is possible that sequences in the homeodomain could
directly
influence dimerization or that the amino acid differences in
the
homeodomain might have a long-range influence on the conformation
of the specificity region resulting in different interactions
with some
of the chimeric
bW alleles.
Identification of single amino acid positions that influence
specificity.
In earlier work, we noted that sequence comparisons
of chimeric alleles with different specificities allowed the
identification of amino acid residues that were important for
specificity (37). Our expanded collection of chimeric
alleles for bE1/bE2 and the new collection for the
bW1/bW2 alleles provided an opportunity to compile a list of
amino acid positions that are involved in the specificity of
interaction. In particular, crosses between strains carrying chimeric
alleles and strains carrying wild-type alleles identified several pairs
of alleles (e.g., bWx6 and bWx9) whose products
differ in sequence at only one amino acid position but which are found
to confer a difference in specificity when tested against strains
carrying wild-type alleles. The mating behavior of the strains carrying
the bW chimeric alleles, which is believed to reflect the
specificity of the interaction between bE and bW
gene products, is shown in Fig. 3. The sequence alignments for the
amino acids encoded by some of those chimeric alleles and for those
encoded by other chimeras that were found to differ in specificity when
tested with wild-type alleles are shown in Fig.
7. These alignments focus attention on
key positions within the border sequences of the regions defined by the
class II bE and bW alleles and identify eight
amino acid positions that influence specificity. These positions
include those encoded by codons 31 and 79 of bE and by
codons 6, 74, 77, 79, 81, and 82 of bW.
View larger version (27K):
[in this window]
[in a new window]
|
FIG. 7.
Sequence alignments for the products of chimeric alleles
that were found to differ in specificity and in a single amino acid
position when tested against the products of wild-type alleles. Eight
amino acid positions that influence specificity (marked with asterisks)
were identified by sequence alignments of alleles whose products were
found to differ in mating reactions when tested with those of wild-type
alleles. Four bE1/bE2 chimeric alleles allowed the
identification of two amino acid positions that influence specificity.
An additional 12 bW1/bW2 chimeric alleles identified six
positions that determine specificity. The results of mating tests
demonstrating the interactions of strains carrying the bW chimeric
alleles are shown in Fig. 3, and the results of mating tests for all of
the alleles are shown in Table 2. The specificity classes are indicated
on the right, the underlined sequences represent the bE1 or bW1
portions of the products of the chimeric alleles, and the remaining
sequences are from bE2 or bW2.
|
|
It is interesting to note that among the eight positions that influence
specificity, four of the amino acid differences involve
a Tyr residue.
These positions have substitutions of Tyr for either
Arg, His or, in
two cases, Cys. Charged or polar amino acids are
present at one or both
of the positions in six of the eight examples.
Only one of the
positions (bE codon 79) has a substitution of
two hydrophobic residues
(Ile and Phe) and one (bW codon 79) has
a substitution of basic
residues (His for Arg). In addition, a
reversal of charge (Lys or Asp)
was found for one position (bW
codon 77). These comparisons of the
amino acids found at positions
that influence specificity suggest that
charge and polarity may
play an important role in the interaction
between the bE and bW
polypeptides. Furthermore, it is striking that
residues with aromatic
side chain rings, i.e., His, Tyr, and Phe are
prominent within
the list of amino acids at the eight positions.
Overall, these
results indicate that it is possible to use differences
between
chimeric alleles to identify single-amino-acid positions
important
for specificity and to catalog the types of residues at those
positions.
The identification of important amino acid positions within the border
regions was extended by the analysis of additional
chimeric alleles
that were found to have different specificities
when strains with
chimeric alleles were used as testers. An important
feature of these
crosses between strains carrying chimeric alleles
was the
identification of interesting interactions between specific
chimeras
with recombination points in or near the N and C borders
of the
specificity regions. For example, alleles
bEx87 and
bEx89 show opposite specificities when tested with various
bW chimeric
alleles (Table
2). The behavior of these and
other alleles with
adjacent recombination points indicates that
recombination has
occurred in regions that are important for
specificity, i.e.,
the N and C borders. These data can also be used to
identify the
amino acid positions that play a role in the specificity
of interactions
between the products of chimeric alleles. Sequence
alignments
of amino acids encoded by chimeric alleles with specificity
differences
are shown in Fig.
8. These
sequence alignments reveal single-amino-acid
differences at important
positions in the allele products and
provide an additional list of the
types of residues that influence
specificity. As with the eight amino
acid positions identified
in the analysis shown in Fig.
7, the majority
of the residues
are charged or polar and few are hydrophobic. In one
position
(bW codon 9), a clear charge difference is present (Asp versus
Lys). At two other positions, the amino acid differences involve
substitution of a polar or charged residue for a hydrophobic residue
(e.g., bE codon 45 and bW codon 80).
View larger version (27K):
[in this window]
[in a new window]
|
FIG. 8.
Alignments of predicted sequences of the products of
chimeric alleles that were found to differ at a single amino acid and
have different specificities when tested against the products of other
chimeric alleles. The portion of each sequence containing the single
amino acid difference (asterisk) is shown, and the specificity class of
each allele is indicated on the right. The patterns of interactions for
each allele pair are shown in Table 2. Note that alleles
bEx87, bEx89, bEx90, and
bEx92 were all designated class III when assayed with
wild-type b1 and b2 mating partners. Differences
in specificity are revealed in mating assays with strains carrying
other chimeric alleles as shown in Table 2.
|
|
The identification of bE codon 45 as a key position is interesting
because this is the only position which shows an influence
and which is
outside of the border regions previously identified
as containing the
important residues. This finding suggests that
the border regions that
were defined by testing
b1 and
b2 chimeric
alleles with wild-type strains may not be definitive when testing
chimeric alleles against each other. That is, bE position 45 may
have a
residue that is important for specificity only in the context
of the
chimeric alleles. This finding reinforces the idea that
the
identification of the residues in the bE and bW N-terminal
dimerization
domains that are important for specificity is dependent
upon the allele
combinations under investigation.
Possible interactions between the N and C border regions.
Chimeric alleles bEx87 and bEx89 encode products
that differ at a single amino acid position (Fig. 8) and, as shown in
Table 2, have different specificities when tested against chimeric bW alleles with recombination points near the N-terminal
border (e.g., bWx12, bWx19, and bWx36)
and within the C-terminal border (e.g., bWx76,
bWx79, bWx80, and bWx82). Although
this allele pair was the only one to clearly exhibit this phenotype,
this finding suggests that interactions may occur between the N- and
C-terminal border regions. That is, the ability of a difference within
one border (defined by bEx87 and bEx89) to alter
the specificity of interactions with alleles with recombination points
near or in both the N- or C-terminal borders may indicate that the
borders cooperate.
| |
DISCUSSION |
Specificity determinants in the variable N-terminal portions of bE
and bW.
The bE and bW genes of U. maydis each exist in a series of at least 25 alleles that are
primarily distinguished by variability in the regions encoding the
N-terminal 100 to 150 amino acids (12, 17, 24, 28, 29).
Dimerization of bE and bW proteins encoded by alleles from mating
partners of different specificities has been demonstrated
(15) and is thought to establish a transcription factor that
controls morphogenesis and pathogenesis. The bE and bW proteins encoded
by the same strain (like specificity) fail to dimerize (15).
A primary goal in the analysis of the b proteins has been the
identification of regions of bE and bW that control the specificity of
interaction. The N-terminal regions of bE and bW were obvious targets
for this analysis because these regions contain most of the allelic
variation and mediate dimerization (12, 15, 17). In
addition, our previous work on chimeric bE alleles revealed that
recombination within the 5' proximal coding regions resulted in alleles
with novel specificity (37).
The construction and analysis of dual sets of chimeric alleles for
bE1/bE2 and
bW1/bW2 that are described here
provided an
opportunity to further refine our view of the N-terminal
specificity
regions believed to control the recognition between bE and
bW
proteins. Previously, we found that
bE1/bE2 chimeras that
contained
recombination points in the central portion of the variable
region
were of particular interest because they had specificities
different
from either parental allele (
37). These alleles
were designated
class II to distinguish them from alleles that had not
changed
specificity (class I) or that had switched from one parental
specificity
to the other (class III). One major finding from the
extension
of our work to include chimeras of
bW1 and
bW2 is that the same
three classes of alleles could be
identified, including the class
II group with novel specificity.
However, in contrast to the findings
for
bE1 and
bE2, the transition between class II and class III
chimeras
was not distinct for
bW1 and
bW2. Rather, a
pattern of
switching between class II and class III specificities was
observed.
This result served to focus attention on the borders of the
region
defined by the class II alleles and provided the framework for
more detailed studies to identify sequences that control specificity.
Throughout our analysis of the chimeric alleles, we have made the
assumption that the differences in specificity indicated
by the
presence or absence of filamentous cell growth in mating
tests reflect
differences in dimerization ability between bE and
bW polypeptides.
This assumption is based on the demonstrated
correlation between
dimerization and mating specificity reported
by Kämper et al.
(
15) for the bE and bW proteins. That is,
Kämper et
al. (
15) have employed in vitro and in vivo assays
to show
that the N-terminal variable portions of the b proteins
mediate
dimerization and that these same regions control mating
specificity. A
similar relationship has also been reported for
analogous
homeodomain-containing mating type proteins in
Schizophyllum commune and
Coprinus cinereus (
1,
36,
38).
It is formally
possible, however, that other explanations account for
the different
activities of the chimeric bE and bW proteins analyzed in
our
study. For example, there may be differences in levels or stability
of the chimeric proteins compared to the parental proteins.
Furthermore,
it is possible that a negative mating test may reflect a
difference
in the activity of a heterodimer (e.g., failure to act as a
transcriptional
repressor or activator) rather than a failure to
dimerize. Although
these other possibilities must be kept in mind, we
would note
that the chimeric alleles that we have constructed have been
used
to directly replace the parental alleles by homologous
recombination.
Thus, the chromosomal location is the same for the
chimeric and
parental alleles, and we would expect transcription of the
genes
to be identical. Also, the chimeric alleles were constructed such
that deletions or insertions did not occur at the site of
recombination.
That is, the chimeric alleles represent genes with novel
combinations
of parental sequences rather than mutated versions. Given
these
considerations and the description of Kämper et al.
(
15) of
amino acid changes in the N-terminal regions that
influence both
dimerization and mating specificity, we favor the
interpretation
that our mating tests reflect differences in the
abilities of
chimeric bE and bW proteins to dimerize. The discussion
below
is presented with this interpretation in mind.
In previous work, we explored the mating behavior of the strains
carrying the class II chimeric alleles of
bE1 and
bE2 to
gain insight into the novel specificities of these
alleles (
37).
For example, we performed mating reactions
between strains carrying
class II alleles of
bE1/bE2 and
strains with wild-type alleles
different from
b1 and
b2 (
bD,
bI, and
bM) to
demonstrate that
the class II alleles are not simply constitutively
active with
any other bW protein (
37). That is,
recombination within the
specificity region does not simply result in
constitutive compatibility
because the mating reactions failed with
bI and
bM strains. In
addition, we performed the
mating tests with the strains with
class II alleles and strains with
three additional
b specificities
(
bH,
bJ, and
bL) to search for specificity differences
between
class II
bE alleles. In these experiments, we found
that all three
class II alleles had the same specificity. Not
surprisingly, we
also found that strains of opposite
a
mating type that carried
different class II alleles of
bE1/bE2 failed to mate with each
other. This was expected
because of the genetic evidence indicating
that specificity is
determined by the interactions between bE
and bW polypeptides
(
12). As described below, the construction
of a set of
bW1/bW2 chimeric alleles provided an opportunity to
retest
the class II alleles of bE1/bE2 for differences in specificity.
Crosses between chimeric alleles identify short regions containing
specificity determinants.
The availability of dual sets of
chimeric alleles of bE1/bE2 and bW1/bW2 allowed
crosses to be performed between all combinations of alleles
representing the three specificity classes (Table 2). The basic finding
from this work was that the products of class II alleles of
bE generally fail to interact in a compatible manner with
the products of class II alleles of bW. Our interpretation of this result is that the borders of the region defined by class II
alleles contain the important determinants of recognition and that
artificial combinations of these borders (resulting from recombination
in the intervening region) generate alleles with novel specificities.
In this context, the 40-amino-acid region defined by the class II
bE1/bE2 alleles and the analogous 70-amino-acid region for
the bW1/bW2 alleles may represent the intervals between the
borders that influence specificity. For bE1 and bE2, these specificity
borders are encoded by codons 31 to 39 and by codons 79 to 92; for bW1
and bW2, these sequences are encoded by codons 2 to 9 and 74 to 83.
As shown in Fig.
6, we have designated the border regions (10 to 20 amino acids) defined by the analysis of class II alleles
as the N and C
sequences. These sequences have specificities N1
and C1 for bE1 and
bW1, and N2 and C2 for bE2 and bW2. In a self
interaction (e.g., bE1
with bW1), regions of like specificity
(N1 with N1 and C1 with C1)
would prevent formation of the heterodimer.
In a nonself interaction
(e.g., N1 with N2 and C1 with C2), the
specificity regions would allow
dimerization. Chimeric proteins
encoded by class II alleles would fail
to interact with each other
because these products would have like
specificity borders (Fig.
6). That is, N1 would interact with N1 and C2
would interact with
C2, resulting in a situation similar to that
occurring with the
bE and bW products of wild-type self alleles. The
idea that the
N and C regions contain important residues for
specificity is
supported by the finding that the bW1 and bW2 chimeras
did not
show a distinct C-terminal transition between alleles with
class
II specificity and alleles with class III specificity. Instead,
recombination events with the C sequence resulted in alleles that
showed a pattern of alternating specificities (Fig.
2B). In addition,
the amino acid positions that influence specificity, as identified
by
comparisons of chimeric alleles, are located mainly in the
N and C
regions (see below).
The analysis of an additional set of chimeric alleles between
bW1 and
bW3 (Fig.
4) supports the importance of
the N and C
borders defined for the
b1 and
b2
genes. That is, the construction
of chimeric alleles for the
bW1 and
bW3 alleles confirmed the
presence of an
N-terminal specificity sequence encoded by the
first 10 codons.
Interestingly, the construction and analysis
of chimeric alleles from
bW1 and
bW3 indicates that some naturally
occurring alleles have products that differ at only one of the
N or C
regions (e.g., N1) and that sequence differences in one
region are
sufficient to provide a different specificity. In terms
of the
specificity borders,
bW3 appears to be a naturally occurring
chimera whose product has an N3 and C1 combination of borders
(Fig.
2A). This suggests that new specificities could be generated
via
recombination between different alleles to reassort the N
and C
sequences; this type of event is demonstrated by the nonparental
specificity of class II chimeric alleles.
Identification of amino acid positions important for the
specificity of recognition.
The sequence comparisons of pairs of
chimeric alleles that differ in specificity and that are neighbors on
the specificity maps (Fig. 2) provided a means of identifying amino
acid positions that influence specificity. Initially, eight of these
positions were identified through crosses between strains carrying
chimeric alleles and strains carrying wild-type b1 or
b2 sequences (Fig. 7). In general, most of the amino acids
found at the eight positions were either charged or polar and
relatively few hydrophobic residues were present. An inspection of the
types of residues occupying the eight positions revealed a
preponderance (six positions) of aromatic amino acids (His, Phe, or
Tyr). Although it is difficult to draw definite conclusions about the
role of aromatic amino acids, it is interesting to note that these
types of amino acids have been found to play important roles in
antigen-antibody binding (8, 23, 25, 33).
An additional eight amino acid positions that influence specificity
were identified from crosses between strains that each
carry chimeric
alleles (Fig.
8). In these amino acid positions,
the majority of
residues were polar or charged, but only one residue
had an aromatic
side chain ring (His), and Tyr and Phe were not
found. We speculate
that the preponderance of aromatic amino acids
found in the first set
of eight positions, compared with the second
set, may reflect
differences in the interactions of the products
of chimeric alleles
with wild-type products compared with interactions
between chimeric
proteins. Taken as a group, the 16 pairs of the
alternate residues (32 amino acids) present at the key positions
reflect the preponderance of
polar and charged residues; that
is, 24 of 32 residues were polar or
charged, 7 of 32 residues
were hydrophobic, and 1 was Gly.
Overall, the data from Table
2 and Fig.
7 and
8 identified six
positions for bE1/bE2 (codons 31, 45, 79, 87, 89, and 90)
and 10 positions for bW1/bW2 (codons 2, 6, 9, 74, 76, 77, 79,
80, 81, and 82)
that influence specificity. It is noteworthy that
these positions are
all found in the N and C border regions except
position 45 of bE. Thus,
the locations of the key amino acids
reinforce the idea that the
failure of class II bE and bW chimeric
alleles to interact results from
the presence of self combinations
of borders (Fig.
4A). Given that
chimeric alleles were constructed
for only 50 to 60% of the potential
positions for
bE and
bW, it
is possible that
additional positions are important in the interactions
of these allele
pairs. However, for
bW1 and
bW2, chimeric alleles
were obtained for 4 of 6 potential positions (positions 2, 6,
9, 12) in
the N-terminal region (codons 1 to 15) and for 10 of
11 sites
(positions 72, 73, 74, 76, 77, 79, 80, 81, 82, 83) in
the C-terminal
region (codons 70 to 85). For
bE1 and
bE2,
chimeric
alleles were constructed for all three potential N-terminal
sites
(positions 28, 31, and 39) between positions 25 and 40 and for
all five potential C-terminal sites (positions 79, 80, 81, 82,
and 83)
between positions 75 and 90. Thus, the majority of potential
chimeric
alleles have been constructed for the N and C regions
and the majority
of the important amino acid positions have probably
been identified for
these allele pairs.
Kämper et al. (
15) have shown that the variable
N-terminal regions of bE and bW control dimerization such that
heterodimers
arise from polypeptides with different specificities
(e.g., bE1
with bW2) but not from polypeptides with like specificities
(e.g.,
bE1 with bW1). In addition, mutations that allowed interaction
between the self polypeptide combination of bE2 and bW2 were
identified.
In general, these mutations were found to increase
hydrophobicity,
and it was suggested that the wild-type residues
involved were
important for the failure of self combinations to
interact. Two
additional mutations resulted in a change in charge,
implying
a contribution from polar interactions for dimerization.
Combining
the results from this work with the analysis of chimeric
alleles
leads to the general idea that a number of key amino acid
positions
control recognition by influencing dimerization. In general,
however,
the amino acid changes described by Kämper et al. that
resulted
in an increase in hydrophobicity promoted interaction
(dimerization)
between the self combination of bE2 and bW2 polypeptides
(
15).
In contrast, the positions identified for chimeric
alleles suggest
a prominent role for charged or polar residues,
including aromatic
amino acids. These differences may reflect the fact
that the substitutions
identified by Kämper et al. represented
changes that allowed
self interaction. In the case of chimeric alleles,
the interactions
of novel combinations of self and nonself sequences
were explored.
Chimeric alleles for other homeodomain mating proteins in
fungi.
Homeodomain proteins encoded by multiallelic genes and
having roles in sexual development have also been characterized for the
mushroom fungi C. cinereus and S. commune. These
proteins, designated HD1 and HD2 for C. cinereus
(1) and Y and Z for S. commune, also contain the
determinants of allelic specificity in N-terminal regions, as revealed
by chimeric allele analysis. In the case of the HD1 and HD2 proteins of
C. cinereus, specificity is determined by the N-terminal 160 to 170 amino acids (1). For S. commune Z
proteins, seven chimeric alleles were constructed between Z4 and Z5,
and these defined a specificity region between codons 19 and 60 (36). Eight chimeric alleles for Y4 and Y3 were also
constructed, and a region determining specificity was found between
codons 17 and 72 (38). As with the class II chimeric alleles
of bE1 and bE2 (37), Y4/Y3 chimeric
genes with exchange points between codons 17 and 72 had specificities
different from either parental allele. This result indicates that in
the case of the Y alleles of S. commune, the borders of the
region defined by alleles with novel specificities carry the important
determinants of recognition. A similar situation would probably be
revealed by chimeric alleles with recombination in the region between
positions 19 and 60 of the Z proteins. Overall, these results suggest
that a common mechanism and perhaps a common structural organization may be employed to determine self versus nonself recognition for the
homeodomain-containing mating-type proteins in basidiomycetes.
The use of chimeric proteins to study recognition.
A chimeric
strategy for identifying specificity determinants has been employed in
other systems involving recognition between polypeptides. For example,
the dimerization specificity of the bacteriophage P22 repressor has
been studied by making chimeras between P22 and homologous repressor
protein 434 (10). In addition, an attempt to determine the
basis of multiallelic self-incompatibility in plants was carried out by
exchanging domains between allelic S-RNases from Nicotiana
alata (39). Chimeric proteins have also been used to
study protein-protein interactions during ligand-receptor recognition
(7, 20, 21, 32). In fact, our observation that bE
and bW class II chimeric alleles have specificities
different from either parent is not unique to fungal mating-type
systems; a similar phenomenon has recently been reported for chimeras
of two glycoprotein hormones (4). Specifically, the chimeras
of human chorionic gonadotropin (hCG) and human follitropin (hFSH) were
shown to exhibit activity unique to a third family member, human
thyrotropin (hTSH). This result was explained by a model stating that
the specificity between ligand and receptor was mediated by
"inhibitory determinants" that restricted binding to only the appropriate combinations (4, 22). The construction of
chimeras was thought to disrupt the inhibitory determinants and unmask activities characteristic of other members of the protein family.
It is interesting to speculate that an inhibitory determinant model
such as that described for receptor-ligand interactions
may be
applicable to the problem of specificity determination
at the
multiallelic
b locus. In the case of
b genes,
specificity
may result from interactions that prevent dimerization
between
bE and bW proteins derived from the same strain. That is, there
may be amino acid residues positioned to interfere with dimerization
between bE1 and bW1, and these inhibitory determinants may be
positioned differently for each self allele combination. Thus
a set of
interfering residues could prevent dimerization between
self allele
combinations of bE and bW; presumably, the residues
would not directly
oppose each other for nonself allele combinations.
The amino acid
residues in the borders defined by chimeric allele
analysis may
represent inhibitory determinants. Site-directed
mutagenesis and in
vitro protein interaction studies, combined
with access to the
three-dimensional structure of the bE and bW
proteins, will be needed
to explore this possibility.
| |
ACKNOWLEDGMENTS |
This work was supported by Research and Collaborative Project
grants from NSERC (to J.W.K.) and by postgraduate fellowships from
NSERC and the Killam Foundation (to A.R.Y.).
We thank Jeanette Johnson-Beatty, George Athwal, and Alan Au for
technical assistance and Lisa Gentile and Lawrence McIntosh for
comments on the manuscript.
| |
FOOTNOTES |
*
Corresponding author. Mailing address: Biotechnology
Laboratory, 237-6174 University Blvd., University of British Columbia, Vancouver, B.C., Canada V6T 1Z3. Phone: 604-822-4732. Fax:
604-822-6097. E-mail: kronstad{at}unixg.ubc.ca.
| |
REFERENCES |
| 1.
|
Banham, A. H.,
R. N. Asante-Owusu,
B. Gottgens,
S. A. J. Thompson,
C. S. Kingsnorth,
E. J. C. Mellor, and L. A. Casselton.
1995.
An N-terminal dimerization domain permits homeodomain proteins to choose compatible partners and initiate sexual development in the mushroom Coprinus cinereus.
Plant Cell
7:773-783[Abstract].
|
| 2.
|
Banuett, F.
1995.
Genetics of Ustilago maydis, a fungal pathogen that induces tumors in maize.
Annu. Rev. Genet.
29:179-208[Medline].
|
| 3.
|
Bölker, M.,
M. Urban, and R. Kahmann.
1992.
The a mating type locus of U. maydis specifies cell signalling components.
Cell
68:441-450[Medline].
|
| 4.
|
Campbell, R. K.,
E. R. Bergert,
Y. Wang,
J. C. Morris, and W. R. Moyle.
1997.
Chimeric proteins can exceed the sum of their parts: implications for evolution and protein design.
Nat. Biotechnol.
15:439-443.
[Medline] |
| 5.
|
Cease, K. B.,
C. A. Potcova,
C. J. Lohff, and M. E. Zeigler.
1994.
Optimized PCR using Vent polymerase.
PCR Methods Appl.
3:298-300[Medline].
|
| 6.
|
Christensen, J. J.
1963.
.
Corn smut caused by Ustilago maydis. Monograph no. 2.
American Phytopathological Society, Saint Paul, Minn.
|
| 7.
|
Cunningham, B. C.,
P. Jhurani,
P. Ng, and J. A. Wells.
1989.
Receptor and antibody epitopes in human growth hormone identified by homolog-scanning mutagenesis.
Science
243:1330-1336[Abstract/Free Full Text].
|
| 8.
|
Davies, D. R., and G. H. Cohen.
1996.
Interactions of protein antigens with antibodies.
Proc. Natl. Acad. Sci. USA
93:7-12[Abstract/Free Full Text].
|
| 9.
|
Davies, D. R.,
E. A. Padlan, and S. Sheriff.
1990.
Antibody-antigen complexes.
Annu. Rev. Biochem.
59:439-473[Medline].
|
| 10.
|
Donner, A. L.,
P. A. Carlson, and G. B. Koudelka.
1997.
Dimerization specificity of P22 and 434 repressors is determined by multiple polypeptide segments.
J. Bacteriol.
179:1253-1261[Abstract/Free Full Text].
|
| 11.
|
Froeliger, E. H., and S. A. Leong.
1991.
The a mating-type alleles of Ustilago maydis are idiomorphs.
Gene
100:113-122[Medline].
|
| 12.
|
Gillissen, B.,
J. Bergemann,
C. Sandmann,
B. Schroer,
M. Bölker, and R. Kahmann.
1992.
A two-component system for self/non-self recognition in Ustilago maydis.
Cell
68:647-657[Medline].
|
| 13.
|
Holliday, R.
1974.
The genetics of Ustilago maydis, p. 575-595. In
R. C. King (ed.), Handbook of genetics, vol. 1.
Plenum Press, New York, N.Y.
|
| 14.
|
Kahmann, R., and M. Bölker.
1996.
Self/nonself recognition in fungi: old mysteries and simple solutions.
Cell
85:145-148[Medline].
|
| 15.
|
Kämper, J.,
M. Reichmann,
T. Romeis,
M. Bölker, and R. Kahmann.
1995.
Multiallelic recognition: nonself-dependent dimerization of the bE and bW homeodomain proteins in Ustilago maydis.
Cell
81:73-83[Medline].
|
| 16.
|
Kronstad, J. W., and S. A. Leong.
1989.
Isolation of two alleles of the b locus of Ustilago maydis.
Proc. Natl. Acad. Sci. USA
86:978-982[Abstract/Free Full Text].
|
| 17.
|
Kronstad, J. W., and S. A. Leong.
1990.
The b mating-type locus of Ustilago maydis contains variable and constant regions.
Genes Dev.
4:1384-1395[Abstract/Free Full Text].
|
| 18.
|
Kronstad, J. W.
1996.
Pathogenesis and sexual development of the smut fungi, p. 141-186. In
G. Stacey, and N. T. Keen (ed.), Plant-microbe interactions, vol. 1.
Chapman and Hall, New York, N.Y.
|
| 19.
|
Lasky, L. A.
1995.
From sticky zippers to morphology.
Nat. Struct. Biol.
2:258-261[Medline].
|
| 20.
|
Leeb, T.,
S. A. Mathis, and L. M. F. Leeb-Lundberg.
1997.
The sixth transmembrane domains of the human B1 and B2 bradykinin receptors are structurally compatible and involved in discriminating between subtype-selective agonists.
J. Biol. Chem.
272:311-317[Abstract/Free Full Text].
|
| 21.
|
Maden, M.
1996.
Retinoids in patterning: chimeras win by a knockout.
Curr. Biol.
6:790-793[Medline].
|
| 22.
|
Moyle, W. R.,
R. K. Campbell,
R. V. Myers,
M. P. Bernard,
Y. Han, and X. Wang.
1994.
Co-evolution of ligand-receptor pairs.
Nature
368:251-255[Medline].
|
| 23.
|
Padlan, E. A.
1990.
On the nature of antibody combining sites: unusual structural features that may confer on these sites an enhanced capacity for binding ligands.
Proteins
7:112-124[Medline].
|
| 24.
|
Puhalla, J. E.
1970.
Genetic studies of the b incompatibility locus of Ustilago maydis.
Genet. Res.
16:229-232.
|
| 25.
|
Rini, J. M.,
U. Schulze-Gahmen, and I. A. Wilson.
1992.
Structural evidence for induced fit as a mechanism for antibody-antigen recognition.
Science
255:959-965[Abstract/Free Full Text].
|
| 26.
|
Sambrook, J.,
E. F. Fritsch, and T. Maniatis.
1989.
.
Molecular cloning: a laboratory manual, 2nd ed.
Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.
|
| 27.
|
Sarkar, G., and S. S. Sommer.
1990.
The megaprimer method of site-directed mutagenesis.
Biotechniques
8:404-407[Medline].
|
| 28.
|
Schulz, B.,
F. Banuett,
M. Dahl,
R. Schlesinger,
W. Schäfer,
T. Martin,
I. Herskowitz, and R. Kahmann.
1990.
The b alleles of U. maydis, whose combinations program pathogenic development, code for polypeptides containing a homeodomain-related motif.
Cell
60:295-306[Medline].
|
| 29.
|
Silva, J.
1972.
Alleles at the b incompatibility locus in Polish and North American populations of Ustilago maydis (DC) Corda.
Physiol. Plant Pathol.
2:333-337.
|
| 30.
|
Specht, C. A.,
A. M. Munoz-Rivas,
C. P. Novotny, and R. C. Ullrich.
1991.
Transformation of Schizophyllum commune: an analysis of specific properties.
Exp. Mycol.
15:326-335.
|
| 31.
|
Spellig, T.,
M. Bölker,
F. Lottspeich,
R. W. Frank, and R. Kahmann.
1994.
Pheromones trigger filamentous growth in Ustilago maydis.
EMBO J.
13:1620-1627[Medline].
|
| 32.
|
Tian, Y.,
L. Wu,
D. L. Oxender, and F. Chung.
1996.
The unpredicted high affinities of a large number of naturally occurring tachykinins for chimeric NK1/NK3 receptors suggest a role for an inhibitory domain in determining receptor specificity.
J. Biol. Chem.
271:20250-20257[Abstract/Free Full Text].
|
| 33.
|
Tsumoto, K.,
K. Ogasahara,
Y. Ueda,
K. Watanabe,
K. Yutani, and I. Kumagai.
1995.
Role of Tyr residues in the contact region of anti-lysozyme monoclonal antibody HyHEL10 for antigen binding.
J. Biol. Chem.
270:18551-18557[Abstract/Free Full Text].
|
| 34.
|
Wang, J.,
D. W. Holden, and S. A. Leong.
1988.
Gene transfer system for the phytopathogenic fungus Ustilago maydis.
Proc. Natl. Acad. Sci. USA
85:865-869[Abstract/Free Full Text].
|
| 35.
|
Wells, J. A.
1996.
Binding in the growth hormone receptor complex.
Proc. Natl. Acad. Sci. USA
93:1-6[Abstract/Free Full Text].
|
| 36.
|
Wu, J.,
R. C. Ullrich, and C. P. Novotny.
1996.
Regions in the Z5 mating gene of Schizophyllum commune involved in Y-Z binding and recognition.
Mol. Gen. Genet.
252:739-745[Medline].
|
| 37.
|
Yee, A. R., and J. W. Kronstad.
1993.
Construction of chimeric alleles with altered specificity at the b incompatibility locus of Ustilago maydis.
Proc. Natl. Acad. Sci. USA
90:664-668[Abstract/Free Full Text].
|
| 38.
|
Yue, C.,
M. Osier,
C. P. Novotny, and R. C. Ullrich.
1997.
The specificity determinant of the Y mating-type proteins of Schizophyllum commune is also essential for Y-Z protein binding.
Genetics
145:253-260[Abstract].
|
| 39.
|
Zurek, D. M.,
B. Mou,
B. Beecher, and B. McClure.
1997.
Exchanging sequence domains between S-RNases from Nicotiana alata disrupts pollen recognition.
Plant J.
11:797-808[Medline].
|
Mol Cell Biol, January 1998, p. 221-232, Vol. 18, No. 1
0270-7306/98/$04.00+0
Copyright © 1998, American Society for Microbiology. All rights reserved.
This article has been cited by other articles:
-
Chew, E., Aweiss, Y., Lu, C.-y., Banuett, F.
(2008). Fuz1, a MYND domain protein, is required for cell morphogenesis in Ustilago maydis. Mycologia
100: 31-46
[Abstract]
[Full Text]
-
Martinez, C., Roux, C., Jauneau, A., Dargent, R.
(2002). The biological cycle of Sporisorium reilianum f.sp. zeae: an overview using microscopy. Mycologia
94: 505-514
[Abstract]
[Full Text]
-
Wu, J., Glass, N. L.
(2001). Identification of Specificity Determinants and Generation of Alleles with Novel Specificity at the het-c Heterokaryon Incompatibility Locus of Neurospora crassa. Mol. Cell. Biol.
21: 1045-1057
[Abstract]
[Full Text]
-
Kues, U.
(2000). Life History and Developmental Processes in the Basidiomycete Coprinus cinereus. Microbiol. Mol. Biol. Rev.
64: 316-353
[Abstract]
[Full Text]
-
May, G., Shaw, F., Badrane, H., Vekemans, X.
(1999). The signature of balancing selection: Fungal mating compatibility gene evolution. Proc. Natl. Acad. Sci. USA
96: 9172-9177
[Abstract]
[Full Text]
Top
Abstract
Introduction
