Cis-dominant regulatory mutations affecting the formation of glucose-repressible alcohol dehydrogenase (ADHII) in Saccharomyces cerevisiae

Summary The formation of ADHII in Saccharomyces cerevisiae is regulated by carbon catabolite repression. There are two genes involved in the formation of ADHII: ADR2, the structural gene as identified by electrophoretic variants and ADR1, possibly a regulatory gene. A new genetic element involved in the regulation of ADHII was identified by three allelic mutants insensitive to strong glucose repression. They were called ADR3 ^c (wild type designation ADR3) and found to be tightly linked to the structural gene, ADR2. The alcohol dehydrogenase found in ADR3 ^c mutants could not be distinguished electrophoretically from the ADHII of the glucose-sensitive wild type, ADR3.Dominance relations between ADR3 ^c and ADR3 were established in diploids heterozygous for ADR3 and the two alleles of ADR2 (ADR2-S: slow ADHII, ADR2-F: fast ADHII). During growth on 10% glucose, an ADR3 ^c ADR2-F/ADR3 ADR2-S heterozygous diploid formed only the fast ADHII variant whereas an ADR3 ^c ADR2-S/ADR3 ADR2-F heterozygote produced only the slow form. This was taken as evidence of the cis-dominance of all ADR3 ^c alleles. The cis-effect of ADR3 ^c was also demonstrated in glucose-derepressed diploids. The ADR3 ^c mutations do not only cause glucose-insensitive ADHII formation, but also reduce the activity of the adjacent structural gene during derepression. Thus ADR3 ^c alleles were considered to be controlling site mutations. No pleiotropic effects were observed on the formation of enzymes related to the function of ADHII.An adr1 ADR2 ADR3 single mutant did not form ADHII. In contrast to this, an adr1 ADR2 ADR3 ^c double mutant formed ADHII at a similar level as an ADR1 ADR2 ADR3 ^c mutant. This showed that ADR3 ^c was epistatic over adr1 (previously suggested as a positive regulatory gene). From this it was concluded that ADR1 is in fact a positive regulatory gene the function of which is required for the expression of the structural gene for ADHII, ADR2. ADR3 is the controlling site for the structural gene ADR2. Mutations at this site, ADR3 ^c , alleviate the requirement for the ADR2 gene product. ADR3 ^c is discussed as a promotor or operator site.     

Isolation and characterization of yeast mutants defective in intermediary carbon metabolism and in carbon catabolite derepression

Summary Yeast mutants deficient in the constitutive ADHI (adc1) were used for the isolation of mutants with deficiencies of the intermediary carbon metabolism, and of mutants defective in carbon catabolite derepression. Mutants were recognized by their inability to grow on YEP-glycerol and/or on ethanol synthetic complete medium. They were either defective in isocitrate lyase (icl1), succinate dehydrogenase (sdh1), or malate dehydrogenase (mdh1, mdh2), mdh-mutants could not uniformely be appointed to one of the known MDH isozymes. Homozygous mdh and sdh1 diploids are unable to sporulate.Three gene loci could be identified by mutants pleiotropically defective in many or all of the enzymes tested. In ccr1 mutants, derepression of isocitrate lyase, fructose-1,6-diphosphatase, ADHII and possibly of the cytoplasmic MDH is prevented, whereas the mitochondrial TCA-cycle enzymes, succinate dehydrogenase and malate dehydrogenase, are not significantly affected. CCR2 and CCR3 have quite similar action spectra. Both genes are obviously necessary for derepression of all enzymes tested. It could be shown that ccr1, ccr2 and ccr3 mutants are not respiratory deficient.     

A positive regulatory gene is required for accumulation of the functional messenger RNA for the glucose-repressible alcohol dehydrogenase from Saccharomyces cerevisiae

The amount of glucose-repressible alcohol dehydrogenase is regulated by the amount of its functional messenger RNA. ADHII² protein was detected by a radioimmune assay and differentiated from ADHI, the classical ADH isozyme, by limited proteolysis with Staphylococcus aureus protease. When yeast containing the wild-type alleles for ADR2 (the ADH II structural locus) and for ADR1 (its positive regulatory gene) were pulse-labeled with [³⁵S]methionine during derepression, radioactive label accumulated in the antibody-precipitated ADHII coterminously with the appearance of ADHII activity. The kinetics of functional ADHII mRNA appearance during derepression in this strain were shown to be the same as those for ADHII protein synthesis in vivo when RNA, extracted from derepressed cells, was translated in a wheat germ cell-free translation system.The role of the positive regulatory gene, ADR1, in ADHII expression was analyzed using two strains mutated at that locus. Yeast containing the adr1-1 allele are incapable of derepressing ADHII activity. When this strain was pulselabeled with [³⁵S]methionine during derepression, approximately one-tenth to one-twentieth the level of ADHII protein synthesis was detected as in the wild-type strain. When RNA was extracted during derepression from cells containing the udr1-1 allele and translated in a wheat germ cell-free system, little functional ADHII mRNA was found to be present.The role of the ADR1 gene was further analyzed using a strain containing the ADR1-5^c allele, which allows constitutive synthesis of ADHII activity. In this strain during glucose repression. ADHII protein synthesis and amount of functional mRNA were at levels comparable to those found for the wild-type strain after complete derepression. Similar kinetics of ADHII protein synthesis and of mRNA accumulation during derepression were observed in the strain carrying the ADR1-5^c allele when compared to that carrying the ADR1 allele, but the absolute amounts were greater by three- to fourfold in cells containing the ADR1-5^c allele. These results indicate that the ADR1 gene acts to increase the level of functional ADHII mRNA during derepression.     

Identification of New Genes Involved in the Regulation of Yeast Alcohol Dehydrogenase II

Recessive mutations in two negative control elements, CRE1 and CRE2, have been obtained that allow the glucose-repressible alcohol dehydrogenase (ADHII) of yeast to escape repression by glucose. Both the cre1 and cre2 alleles affected ADHII synthesis irrespective of the allele of the positive effector, ADR1. However, for complete derepression of ADHII synthesis, a wild-type ADR1 gene was required. Neither the cre1 nor cre2 alleles affected the expression of several other glucose-repressible enzymes. A third locus, CCR4, was identified by recessive mutations that suppressed the cre1 and cre2 phenotypes. The ccr4 allele blocked the derepression of ADHII and several other glucose-repressible enzymes, indicating that the CCR4 gene is a positive control element. The ccr4 allele had no effect on the repression of ADHII when it was combined with the ADR1-5c allele, whereas the phenotypically similar ccr1 allele, which partially suppresses ADR1-5c, did not suppress the cre1 or cre2 phenotype. Complementation studies also indicated that ccr1 and snf1 are allelic. A model of ADHII regulation is proposed in which both ADR1 and CCR4 are required for ADHII expression. CRE1 and CRE2 negatively control CCR4, whereas CCR1 is required for ADR1 function.     

Analysis of mutations affecting Ty-mediated gene expression in Saccharomyces cerevisiae

Summary Yeast translocatable, Ty, elements can cause constitutive synthesis of the glucose-repressible alcohol dehydrogenase (ADHII) when inserted upstream from the 5 end of the structural gene, ADR2. These insertion mutations, ADR3 ^c, are unstable and give rise to secondary ADHII^– mutations. The majority of such mutants, adr3, can be attributed to excision of the insertion sequence, leaving behind a single copy of the -sequence which occurs as a direct repeat at the ends of the Ty elements. A few adr3 mutants appear to be generated by DNA-rearrangements in the vicinity of the Ty insertion. The occurrence of recessive mutants, tye, which are unlinked to ADR2 indicates that the constitutive expression of ADR2 caused by the Ty insertions requires the function of trans-acting genes. These results support the idea that regulation of Ty-linked ADR2 is actively mediated by the insertion sequence and is probably not due to a mere disruption of the wild-type controlling site.     

The regulation of hexokinase and phosphoglucomutase activity in Aspergillus nidulans.

Summary The levels of glucose-6-phosphate and 6-phosphogluconate dehydrogenase in wildtype cells of Aspergillus nidulans varied with the carbon and nitrogen source. In general, hexokinase activity did not vary with carbon or nitrogen source. The ammonium derepressed mutant amrA1 had only 50% of the wildtype level of hexokinase. Phosphoglucomutase activity was low in wildtype cells grown with nitrate, but high in cells grown with ammonium when glucose was the carbon source. A non-inducible mutant, nirA ^-1, in the regulatory gene for nitrate reductase, had high phosphoglucomutase activity when grown with nitrate or ammonium. A constitutive mutant nirA ^c1, in the regulatory gene for nitrate reductase had low phosphoglucomutase activity when grown with nitrate or ammonium. The mutants nir ^-1 and nirA ^c1 are recessive and semi-dominant respectively for abnormal phosphoglucomutase activity.     

Genetics of carbon catabolite repression in Saccharomyces cerevisiae: genes involved in the derepression process

Summary A recessive mutant cat1-1, wild type CAT1, was isolated in Saccharomyces cerevisiae. It did not grow on glycrrol nor ferment maltose even with fully constitutive, glucose resistant maltase synthesis. It prevented derepression of isocitrate lyase, fructose-1,6-diphosphatase and maltase in a constitutive but glucose sensitive maltase mutant. Derepression of malate dehydrogenase was retarded and slowed down. Sucrose fermentation and invertase synthesis was not affected. Respiration was normal. From this mutant, two reverse mutants were isolated. One was recessive, acted as a suppressor of cat1-1 and was called cat2-1, wild type CAT2; the other was dominant and allelic to CAT1 and designated CAT1-2 ^d. CAT1-2 ^d and cat2-1 caused an earlier derepression of enzymes studied but did not affect the repressed nor the fully derepressed enzyme levels. CAT1-2 ^d and cat2-1 did not show any additive effects. It is proposed that carbon catabolite repression acts in two ways. The direct way represses synthesis of sensitive enzymes, during growth on repressing carbon sources whereas the other way regulates the derepression process. After alleviation of carbon catabolite repression, gene CAT1 becomes active and prevents the activity of CAT2 which functions as a repressor of sensitive enzyme synthesis. The CAT2 gene product has to be eliminated before derepression can actually occur. The time required for this causes a delay in derepression after the depletion of a repressible carbon source. cat1-1 cannot block CAT2 activity and therefore, derepression is blocked. cat2-1 is inactive and derepression can start after carbon catabolite repression has ceased. CAT1-2 ^d is permanently active as a repressor of CAT2 and eliminates the delay in derepression.     

Transposable elements associated with constitutive expression of yeast alcohol dehydrogenase II

The yeast structural gene ADR2, coding for the glucose-repressible alcohol dehydrogenase (ADHII), has been isolated by complementation of function in transformed yeast. The chromosomal DNA from nine yeast strains with cis-dominant constitutive mutations (ADR3^c) has been investigated by restriction enzyme analysis, using the cloned ADR2 DNA as a hybridization probe. Seven mutants appear to have insertions of approximately 5.6 kb near the 5′ end of the ADR2-coding region. Four of these insertions have the same restriction pattern as the yeast transposable element Tyl. Two differ from Tyl by the presence of an additional Hind III site, and a seventh insertion differs from Tyl at a number of restriction sites. All are inserted in the same orientation with respect to the structural gene. A DNA fragment containing the ADR2 gene and adjacent sequences from a constitutive mutant has been cloned and shown by heteroduplex analysis to contain an insertion near the 5′ end of the structural gene. The cloned insertion sequence hybridizes to multiple genomic DNA fragments, indicating that it contains a moderately repetitive sequence. Thus it appears that insertion of a transposable element near the 5′ terminus of the structural gene can produce constitutive expression of a normally glucose-repressed enzyme. Such insertions seem to be the most common way of generating cis-dominant constitutive mutations of ADHII.     

A polymorphic variant of the lactate dehydrogenase B subunit in the rat

A new electrophoretic variant of the lactate dehydrogenase B subunit was found in the erythrocytes of the COP strain of the rat. The location of the band after the electrophoresis suggested a product of the structural gene for the B subunit. Two alleles that regulated the high amount (Ldh-2 ^a )or the low amount (Ldh-2 ^b )of the B subunit were found and segregated in Mendelian fashion. The activity was regulated by the closely linked (<1 cM) regulatory gene Ldr-1.     

Genetics and biochemistry of the peptidoglycan-associated proteins b and c of Escherichia coli K12.

Summary To collect information on synthesis and regulation of the peptidoglycan-associated pore-forming outer membrane proteins b and c, mutants resistant to phages Mel and TuIa were analyzed. Genetic analysis showed three linkage groups, corresponding with the genes tolF (phenotype b^-c⁺), meo A (phenotype b⁺c^-) and ompB (phenotypes b^-c^-, b^- c⁺, b⁺⁺ c^- and b⁺⁺ c^±). It has recently been described that also a b⁺ c^- phenotype can occur in the latter linkage group [Chai, T., Foulds, J., J. Bacteriol. 130, 781–786 (1977)]. Among ompB (b^- c⁺)/meoA (b⁺ c^-) double mutants strains were found with the b⁺ c^- phenotype, showing that ompB is not the structural gene for protein b. Studies on purified proteins b and c showed profound differences between the two proteins with respect to the electrophoretic mobility of fragments obtained by treatment with cyanogen bromide, trypsin and chymotrypsin. The amino acid in position three of the amino-termini of proteins b and c, isolated from isogenic strains, were identified as isoleucine and valine respectively. Both the genetic and biochemical results are consistent with a model recently published [Ichihara, S., Mizushima, S., J. Biochem. (Japan) 83, 1095–1100 (1978)] which predicts that tolF and meoA are the structural genes for the proteins b and c respectively and that ompB is a regulatory gene whose product regulates the levels of both proteins.     

Regulation of the expression of iso 2-cytochrome c gene in S. cerevisiae: cloning of the positive regulatory gene CYP1 and identification of the region of its target sequence on the structural gene CYP3

Summary CYP1 is a trans acting regulatory locus modulating both iso 1- and iso 2-cytochrome c synthesis. Genetical analysis of various mutated alleles has allowed us to identify the gene product as a positive regulatory element.The region of the target sequence of the CYP1 product on the iso 2-cytochrome c structural gene was located by molecular and genetic analysis of two cis acting mutations located at the CYP3 locus: CYP3-36 and CYP3-4, which have been shown to arise from the integration of TY1 elements near the promoter site. Determination of the amount of iso 2-cytochrome c synthesized by strains bearing various genetic constructions, in which the cis acting mutations were associated with different alleles of the CYP1 trans acting locus, showed that TY1 inserted into CYP3-36 extinguishes the activation function due to a mutated overproducer allele CYP1-18, while CYP3-4 amplifies this function. This result identifies at least a part of the target sequence of the CYP1 product within the region separating the two TY1 insertions.To clone the CYP1 gene, we took advantage of the iso 2-cytochrome c overproducer phenotype of the mutated allele CYP1-18, which confers a Lactate⁺ phenotype on an iso 1-cytochrome c-deficient strain. Such a phenotype allowed the isolation of a recombinant plasmid YEpJFM1 carrying the mutated allele, able to complement on lactate medium a lactate ^- recipient strain. The identity of the YEpJFM1 sequence with the chromosomal gene was confirmed by homologous recombination at the CYP1 locus.     

Negative control of the galactose operon in E. coli

Summary Non-inducible mutants have been isolated which synthesize the three galactose enzymes with the basal rate both in the absence and in the presence of inducers. These mutations are closely linked to the lysA gene, as are the constitutive mutations in the regulator gene first described by Buttin (1963).The non-inducible mutants are Gal ^– on EMB gal plates. Revertants to the Gal ⁺ phenotpye are constitutive. Heterozygotes have been prepared at the locus of the regulator gene (galR), abd dominance studies involving the different alleles at this locus have been carried out. The non-inducible mutations are dominant over the wildtype, and this in turn is dominant over constitutive mutations in the galR gene.Starting from the non-inducible mutations, deletions have been isolated, which extend from the galR gene into the lysA gene. These are constitutive.The behavior of the non-inducible mutations and of the deletions are strong arguments for negative control of the galactose operon.     

“The Image” of the Regulatory Gene in Experiments with Drosophila

The mutants referred to as facultative dominant lethals were selected in the progeny of gamma-irradiated Drosophila males. The mutant males were viable and fertile, though their crosses with females of the yellow line yielded no daughters. The mutations obtained differed from the common mutations by (1) extremely varying penetrance of F₁ hybrids from crosses with various lines; (2) the uncertain relationships between the mutant and normal alleles; (3) the different expression in somatic and germ cells; (4) the dependence of the expression on the sex of the parent that was the donor of the mutation; (5) the mass morphosis formation and (6) the frequent reversal to the norm. These mutations are assigned to the regulatory group and their specific expression (see above) can be helpful in identifying regulatory gene mutations. We assume that the specific expression of the mutations studied is related to specific properties of the regulatory genes. These properties are as follows: (1) only one out of two homologous regulatory genes is in an active state, (2) in the haploid chromosome set, the regulatory gene is represented by several alleles (cys-alleles); (3) only one allele ensures the regulatory gene activity.     

Isolation and characterization of two new negative regulatory mutants for nitrate assimilation inChlamydomonas reinhardtii obtained by insertional mutagenesis

Plasmid DNA carrying either the nitrate reductase (NR) gene or the argininosuccinate lyase gene as selectable markers and the correspondingChlamydomonas reinhardtii mutants as recipient strains have been used to isolate regulatory mutants for nitrate assimilation by insertional mutagenesis. Identification of putative regulatory mutants was based on their chlorate sensitivity in the presence of ammonium. Among 8975 transformants, two mutants, N1 and T1, were obtained. Genetic characterization of these mutants indicated that they carry recessive mutations at two different loci, namedNrg1 andNrg2. The mutation in N1 was shown to be linked to the plasmid insertion. Two copies of the nitrate reductase plasmid, one of them truncated, were inserted in the N1 genome in inverse orientation. In addition to the chlorate sensitivity phenotype in the presence of ammonium, these mutants expressed NR, nitrite reductase and nitrate transport activities in ammonium-nitrate media. Kinetic constants for ammonium (¹⁴C-methylammonium) transport, as well as enzymatic activities related to the ammonium-regulated metabolic pathway for xanthine utilization, were not affected in these strains. The data strongly suggest thatNrg1 andNrg2 are regulatory genes which specifically mediate the negative control exerted by ammonium on the nitrate assimilation pathway inC. reinhardtii.     

Analysis of yeast ilv 1 CIS control and domain mutants

Summary We have identified a cis control region specific for the ilv 1 gene of Saccharomyces cerevisiae. Mutants designated ilvl-OP ^cmap in this control region located between arg 6 and ilv 1 and result in increased basal levels and constitutive synthesis of the ilv 1 gene products. Furthermore, ilv 1 mutants have been isolated in three different structural domains indicating that the ilv 1 gene may contain a functional intervening sequence specific for one of the two gene products.     

Regulatory mechanism of the tryptophan operon in Escherichia coli: possible interaction between trpR and trpS gene products

Summary The trpS5 mutation (a mutation in the structural gene for tryptophanyl-tRNA synthetase (TRSase) in E. coli), when present in the genetic background of strain KY913 (HfrH), results in the failure to grow at high temperature (42° C) in a complete medium. The rel (RC relaxed) marker present in this strain was found to be partly responsible for this temperature sensitivity. TRSase in such a strain was rapidly inactivated during growth at 42° C in rich media, but not in minimal media or in the presence of chloramphenicol. A partial derepression of anthranilate synthetase formation took place in the presence of excess tryptophan at growth-restricting temperatures. When some of the trpR mutations (including amber mutations) were combined with trpS5, the resulting double mutants (trpR trpS5) were temperature-insensitive, and TRSase was not inactivated at high temperature, in contrast to the trpR ⁺trpS5 strain. This effect of trpR mutations on temperature sensivity was shown not to be a secondary consequence of the constitutive expression of the trp operon. These findings suggest that the trpR ⁺ product interacts with the TRSase of the trpS5 mutant so as to bring about the growth-dependent inactivation of the enzyme. Furthermore, a special class of trpR mutants was obtained whose constitutivity with respect to the trp operon is manifested only in strains carrying trpS5 (but not trpS ⁺) grown at high temperatures. It is proposed that TRSase participates in repression trrough direct interaction with the product of the trpR gene.     

Nitrogen metabolite repression in Aspergillus nidulans

Summary In Aspergillus nidulans, mutations, designated areA^r, can result in the inability to utilise a wide variety of nitrogen sources including amino acids, purines, amides, nitrate, and nitrite, whilst not affecting growth on ammonium. Other allelic areA mutations, designated areA^d, lead to derepression of one or more activities which are ammonium repressible in wild type (areA⁺) strains, whilst not affecting their inducibility. Various areA mutations exhibit a wide variety of phenotypes: areA^r alleles can be temperature sensitive on some nitrogen sources while not on others, and different alleles can be temperature sensitive for utilisation of different nitrogen sources. areA^d alleles can be derepressed for one ammonium-repressible activity, be normally repressible for another, and lead to abnormally low levels for a third. Once again each areA^d allele has its own highly specific phenotype. The inability of areA^r strains to utilise most nitrogen sources is paralleled by low activities of certain ammonium-repressible enzymes. areA^r mutations appear to be epistatic to some but not all regulatory mutations leading to constitutive synthesis of inducible enzymes and also epistatic to gdhA mutations which lead both to loss of NADP-linked glutamate dehydrogenase and to derepression of ammonium-repressible activities. areA^r mutations do not interfere with repair of a large number of auxotrophies in double mutants. Furthermore, although areA^r mutations prevent utilisation of L-arginine, L-ornithine, and L--amino-n-butyrate as nitrogen sources, they do not prevent the metabolism of these compounds necessary for repairing auxotrophies for proline and isoleucine in the appropriate double mutants. Utilisation of acetamide and most amino acids as carbon or carbon and nitrogen sources is unaffected by areA^r mutations, and areA^r strains are able to utilise acetamide and L-proline (but not other amino acids) as nitrogen sources in the presence of non-catabolite-repressing carbon sources such as L-arabinose, glycerol, melibiose, and lactose. Suppressor mutations, designated creA^d, probably leading to loss of carbon catabolite repression, allow utilisation of acetamide and proline as nitrogen sources in areA^r double mutants in the presence of carbon catabolite-repressing carbon sources. creA^d mutations allow ethanol to serve as a source of acetate for pyruvate dehydrogenaseless (pdhA) strains in the presence of carbon catabolite-repressing carbon sources, whereas pdhA single mutants respond to ethanol as sole carbon source only in the presence of non-carbon catabolite-repressing carbon sources. Specific suppressor mutations, designated amd ^d and prn ^d, allow utilisation of acetamide or proline, respectively, in areA^r double mutants.The areA locus can be interpreted as specifying a protein which is capable of (and in most cases essential for) allowing the synthesis of a number of enzymes of nitrogen metabolism but which cannot function in the presence of ammonium (i.e., as specifying a positive regulatory element which mediates ammonium repression) although the possibility that the areA product also plays a negative regulatory role cannot at present be ruled out.     

Identification of self-incompatibility genotypes of almond by allele-specific PCR analysis

In almond, gametophytic self-incompatibility is controlled by a single multiallelic locus (S-locus). In styles, the products of S-alleles are ribonucleases, the S-RNases. Cultivated almond in California have four predominant S-alleles (S ^a, S ^b, S ^c, S ^d). We previously reported the cDNA cloning of three of these alleles, namely S ^b, S ^c and S ^d. In this paper we report the cloning and DNA sequence analysis of the S ^a allele. The S^a-RNase displays approximately 55% similarity at the amino-acid level with other almond S-RNases (S^b, S^c, and S^d) and this similarity was lower than that observed among the S^b, S^c and S^d-RNases. Using the cDNA sequence, a PCR-based identification system using genomic DNA was developed for each of the S-RNase alleles. Five almond cultivars with known self-incompatibility (SI) geno-types were analyzed. Common sequences among four S-alleles were used to create four primers, which, when used as sets, amplify DNA bands of unique size that corresponded to each of the four almond S-alleles; S ^a (602 bp), S ^b (1083 bp), S ^c (221 bp) and S ^d (343 bp). All PCR products obtained from genomic DNA isolated from the five almond cultivars were cloned and their DNA sequence obtained. The nucleotide sequence of these genomic DNA fragments matched the corresponding S-allele cDNA sequence in every case. The amplified products obtained for the S ^a- and S ^b-alleles were both longer than that expected for the coding region, revealing the presence of an intron of 84 bp in the S ^a-allele and 556 bp in the S ^b-allele. Both introns are present within the site of the hypervariable region common in S-RNases from the Rosaceae family and which may be important for S specificity. The exon portions of the genomic DNA sequences were completely consistent with the cDNA sequence of the corresponding S-allele. A useful application of these primers would be to identify the S-genotype of progeny in a breeding program, new varieties in an almond nursery, or new grower selections at the seedling stage. Received: 21 June 1999 / Accepted: 15 November 1999     

Expression level of the SLG gene is not correlated with the self-incompatibility phenotype in the class II S haplotypes of Brassica oleracea

In Brassica, the S-locus glycoprotein (SLG) gene has been strongly implicated in the self-incompatibility reaction. Several alleles of this locus have been sequenced, and accordingly grouped as class I (corresponding to dominant S-alleles) and class II (recessive). We recently showed that a self-compatible (Sc) line of Brassica oleracea expressed a class II-like SLG (SLG-Sc) gene. Here, we report that the SLG-Sc glycoprotein is electrophoretically and immunochemically very similar to the recessive SLG-S15 glycoprotein, and is similarly expressed in stigmatic papillae. Moreover, by seed yield analysis, we observe that both alleles are associated with a self-compatibility response, in contrast with the other known recessive S haplotypes (S2 and S5). By genomic DNA blot analysis, we show the existence of molecular homologies between the Sc and S15 haplotypes, but demonstrate that they are not identical. On the other hand, we also report that the S2 haplotype expresses very low amounts of SLG glycoproteins, although it exhibits a self-incompatible phenotype. These results strongly question the precise role of the SLG gene in the molecular mechanisms that control the self-incompatibility reaction of Brassica.