Current Genetics (19 84) 8 : 107-114

~

~

~

© Springer-Verlag 1984

Identification of regulatory genes of riboflavin permease and a-glucosidase in the yeast Pichia guilliermondii A. A. Sibirny and G. M. Shavlovsky Institute of Biochemistry, Academy of Sciences of Ukrainian S.S.R., Lvov Branch, 290005 Lvov, USSR

Summary. The method for a positive selection of Pichia guilliermondii yeast mutants which constitutively synthesize riboflavin (RF) permease has been developed. A genetic analysis revealed two regulatory genes of negative action (RFP80, RFP81) and one gene of positive action (RFP82); mutations in these loci determined the ability to synthesize RF permease in the medium without an inducer (~-glucosides). The constitutive mutants with cold-sensitive products of RFP80 and RFP81 genes were isolated. Interallelic complementation within RFPSO locus as well as restoration of the wild (inducible) phenotype in some hybrids between recessive rfp80 mutants and dominant RFP82 C mutants were observed. These data suggest a protein structure of products of identified regulatory loci and a direct interaction of the products of RFP80 and RFP82 genes. A meiotic segregants unable to synthesize RF permease in the inducer-containing media (genotype rfp82) were isolated by means of intragenic recombination in RFP82 locus. Epistasis-hypostasis test showed that gene RFP82 acted after gene RFP80. RFP80, RFP81 and RFP82 loci are involved in regulation of biosynthesis of both RF permease and ~-glucosidase. The model for action of RFPSO and RFP82 gene products in the expression of RF permease and ~-glucosidase structural genes ofP. gulliermondii is presented. Key words: Pichia guilliermondii - Regulatory mutants - Riboflavin transport - e-Glucosidase

Introduction The wild strains and RF deficient mutants of the yeast Pichia guilliermondii are incapable of mediated RF

Offprint requests to. A. A. Sibirny

transport from the medium (Sibirny et al. 1977a, c). However, RF deficient mutant MS1-3 was selected growing in the medium with very low RF concentrations which was able to take up and accumulate the great amounts of vitamin B 2 (Shavlovsky et al. 1977; Sibirny et al. 1977 a, d). The study of properties of RF transport system (RF permease) in the strain MS1-3 and in the prototrophs obtained from it showed that this system catalyzed RF uptake against concentration gradient. RF permease was characterized by saturation kinetics, substrate specificity and by dependence on temperature and pH of incubation mixture. RF permease activity was competitively inhibited by glucose and some of its analogs (Sibirny et al. 1978). At the same time glucose was not transported into the cell by means of RF permease (Sibirny et al. 1981). Apparently the revealed transport system has a unique property: it accepts two different classes of compounds (flavins, sugars) but possesses rather a stringent substrate specificity for each class. RF permease proved to be an inducible system. Its synthesis occured only in the presence of ~-glucosides (sucrose, maltose etc). The induction of the system was depressed by translation and transcription inhibitors (Sibirny et al. 1979). e-Glucosides appeared to be inducers of the synthesis not only of RF permease but also that of ~-glucosidase. Some P. guilliermondii mutants were selected which were capable of a constitutive synthesis of RF permease; the formation of ~-glucosidase in such mutants occured consitutitvely too (Sibirny et al. 1979). Here we present a method for positive selection of constitutive mutants for RF permease synthesis as well as results on the genetic analysis of isolated mutants and on the identification of corresponding regulatory genes.

108

A . A . Sibirny and G. M. Shavlovsky: Riboflavin permease and a-glucosidase regulation in yeast

Activeproduct

Inactivecomptex

Activeproduct

Activeproduct

(Inducer, ~z-g[ucosid)e b

Inactivecomplex

Inactivecomplex @@

Inactivecomplex

'

Inactive Inactive ActiveproductActiveproduct (comp[ementati complex ocompl n) ex

nactiveprOdr~Ct

(Inducer)~ ~

Materials and methods Strains. The strains o f genetic line of P. guilliermondii capable of the active RF transport (Rtr +) were used. Selection of such genetic line will be published elsewhere. Strains L13 (mat-, rib2, ade2) and L19 (mat +, rib2, argX) o f the isolated genetic line were used to select regulatory mutants; marker rib2 o f both strains was isogenic.

Genetic analysis o f P. guilliermondii. The conditions for hybridization, sporulation and random spore analysis as well as the composition o f the corresponding media were described earlier (Sibirny et al. 1977b, e). Haploid strains of opposite mating types effectively cross on the acetate medium; the same medium was used for sporulation of hybrids. Since asci o f P. guilliermondii contain mainly two spores, the meiotic segregation was studied by mass spore analysis. The spores were obtained by selective killing o f vegetative diploid cells with 20% ethanol or by separation o f spores from the cells using paraffin oil. The absence of four-spored asci in P. guilliermondii makes a tetrad analysis impossible. Random spore analysis ofP. guilliermondii diploids showed that segregation o f some auxotrophic markers, including certainly monogenic (e.g. ribl), may differ from mendelian (Sibirny et al. 1977b). Apparently it depends on different viability of ascospores of distinct genotypes. This makes difficult a precise determination o f the monogeneity of studied features. At the same time mass spore analysis o f P. guilliermondii does not prevent the identification o f the linkage o f allelic markers (Shavlovsky et al. 1979). Yeast cultivation. The yeast was cultivated in the liquid medium which contained (per 1 1): 20 g o f c a r b o n source; 3 g (NH4)2SO4; 0.5 g KH2PO4; 0.2 g MgSO 4 - 7H20; 0.5 g o f yeast extract (Difco). The meaium contained also growth factors for auxotrophs including RF (from 0.3 to 200 #g/ml depending on the

Fig. I A - E . The model of RFP80 and RFP82 regulatory proteins interaction in the inducible strains and in the different types o f regulatory mutants. For details see text. A inducible strain; B rfp80 constitutive mutant; C RFP82 c constitutive mutant; D rfp82 uninducible mutant; E restoration o f inducibility in some rfpSO x RFP82 c hybrids

requirement o f a represented mutant). The cells were grown on a shaker (200 rpm) at 30 °C during 2 4 - 3 6 h.

Determination of R F permease and a-glucosidase activities. For qualitative testing the uptake of RF by strains studied, they were grown on the solid media supplemented with the high RF concentration (200 /zg/ml) and various carbon sources. Yellow colour o f the biomass on the medium with an inducer (sucrose, 5 g/l) and RF indicates the ability o f the tested strains for the RF uptake (Rtr+); white colottr points to the absence o f RF permease activity (Rtr-). Yellow colour of the biomass in the inducer-free medium (glucose, 5 g/t) and RF, indicates the constitutive synthesis o f RF permease. The intensity o f yellow colour permits a semiquantitative estimation o f the RF permease activity. The quantitative determination of the RF permease activity was carried out by the estimation o f the rate of RF uptake by intact washed cells essentially as described earlier (Sibirny et al. 1979). a-Glueosidase activity was determined from the rate of p-nitrophenyl-a-D-glucopyranozide hydrolysis by digitoninpermeabilized whole cells as described previously (Sibirny et al. 1979).

Selection of constitutive mutants for R F permease synthesis. It is known (Shavlovsky et al. 1977) that the requirement of RF deficient Rtr + strain MS1-3 for exogenous RF depends on the source of carbon nutrition: optimal growth o f this strain in the sucrose medium occurs at significantly lower RF concentrations than in the medium with glucose. In the experiments with strains L13 and L19 o f genetic line, a similar regularity was found. On the solid media with sucrose or maltose (5 g/l) these strains grew well already in the presence o f 0.2 ~g/ml of RF while in the media with glucose (5 g/1 they demanded no less than 2 gg RF/ml. Concentration of tested sugars was low since they are known to inhibit RF permease activity (Sibirny et al. 1978).

A. A. Sibirny and G. M. Shavlovsky: Riboflavin permease and c~-glucosidaseregulation in yeast It was suggested that the selection of RF deficient Rtr+ strains growing in the inducer-free medium (e.g. with glucose) at low (0.2 0.5 ~tg/ml) concentrations of RF will permit the isolation of constitutive mutants for the RF permease synthesis. To test this hypothesis, we carried out the UV mutagenesis of strain LI 3. The irradiated suspension was spread on the plates with glucose (5 g/l), RF (0.5 gg/ml) and adenine (50 gg/ml). The suspension of 48 h yeast culture pregrown on wort-agar slants was UV irradiated by the dose which provided 10% survivors. Approximately 100-200 thousands of viable cells per plate were spread. After 6 days of incubation at 30 °C, approximately 100 colonies per plate appeared. 644 colonies were streaked on the selective medium and thereafter were replica-plated to the medium with glucose (5 g/l) and RF (200 ~g/ml), and to the medium with glucose without RF. It was found that 61 strains (9.5%) formed colourless colonies on the medium with glucose plus RF, and 37 of them (5.7% of total quantity) proved to be RF prototrophs. The others 583 strains (90.5%) formed yellow colonies on the media with RF (200 ~g/ml) and glucose (5 g/l) or any utilizable carbon source tested (sucrose, fructose, and mannitol). Thus these mutants constitutively synthesized RF permease. The constitutive for RF permease synthesis mutants from the strain of opposite mating type, i.e. L19, were selected in similar wise.

Results

Complementation and segregation analysis of constitutive mutants for RF permease synthesis 74 constitutive mutants isolated from strain L13 and 65 constitutive mutants from strain L]9 after a twofold cloning were crossed with initial inducible strains as well as between themselves in all possible combinations. Using a standard for P. guilliermondii hybridization acetate medium it was found out that many pairs o f constitutive mutants did not form diploids. Yet the change of acetate medium for the medium with tomato juice (Sibirny et al. 1977b) permitted the isolation of all possible for this cross hybrids (75 x 66 = 4,950 hybrids). To test the ability of hybrids for a constitutive RF permease synthesis, replica-plating from the medium with tomato juice onto the medium with glucose (5 g/l) plus RF (200 tag/ml) was carried out. This medium served simultaneously for hybrid selection and testing the ability of resulted hybrids to synthesize RF permease in the inducer-free medium (by the colour of hybrid colonies). Simultaneously the colour o f obtained hybrid colonies in the medium with the inducer (sucrose, 5 g/l) and RF (200 ~tg/ml) was determined. It was found that in the latter medium all strains synthesized RF permease. The quantitative determination of RF perincase activity (see below) confirmed the qualitative estimation of phenotypes of studied strains obtained by virtue o f colour test. The ability for constitutive synthesis of RF permease in two mutants, L13-46 and L19-6 (the first was isolated

109

from strain L13, the second from L19), was dominant. The hybrid colonies obtained by crossing of these mutants with initial inducible strains were yellow in the medium with glucose (5 g/l) plus RF (200/Jg/mi), and RF permease activities were approximately the same after growing these hybrids in the media with glucose or sucrose (Table 1). In one mutant, L13-48 (isolated from strain L13), constitutive character of RF permease synthesis was semidominant: the colonies of hybrid L13-48 x L19 were poorly yellow on the medium with glucose plus RF and RF permease activity of this diploid grown in glucose medium was considerably less than that after cultivation in sucrose medium (Table 1). In remaining 136 mutants constitutive phenotype was recessive. The hybrids isolated by crossing of these mutants with initial strains formed colourless colonies on the medium with glucose (5 g/l) plus RF (200/Jg/ml) and practically did not contain RF permease activity when grown in glucose medium (Table 1). Complementation analysis o f recessive mutants showed that all 64 mutants isolated from strain L19 belong to one class of complementation (class I); 54 mutants selected from strains L13 also belong to this class. The hybrids obtained by crossing o f mutants of class I synthesized RF permease constitutively (Table 1). The other 18 constitutive mutants isolated from strain L13 produced inducible hybrids when crossed with mutants of complementation class I. A diploid obtained by hybridization of one such mutant, L13-6, with strain L19 was spread on spomlation medium, and thereafter the random spore progeny of this hybrid was analysed. Three constitutive segregants (genotype: mat +, rib2, argX), T2, T3, T4, were picked out. 18 constitutive mutants not belonging to complementation class I were crossed with segregants T2, T3, T4. Resulted hybrids possessed constitutive RF permease (Table 1); this complementation class was designated as class II. The data obtained suggest the existence o f two regulatory genes, recessive mutations in which resulted in the constitutive synthesis of RF permease. The mutants of complementation class I and class II were designated as rfp80 (riboflavin permease) and rfp81, respectively. Segregation o f constitutive strains from diploids obtained by crossing ofrfp80 and rfpS1 mutants (L13-1 and L13-6, respectively) with initial inducible strain L19 differed from mendelian: the number o f inducible segregants considerably exceeded the number of constitutive ones. The hybrid obtained by crossing o f dominant mutant L13-46 with initial strain L19 segregated almost the equal number o f constitutive and inducible strains, and the second similar hybrid (L13 x L19-6) segregated constitutive rather than inducible strains (Table 2). These data confirm the complications to determine the monogeneity of markers in P. guilliermondii basing on the random spore analysis.

110

A.A. Sibirny and G. M. Shavlovsky: Riboflavin permease and c~-glucosidase regulation in yeast

Table 1. RF permease and c~-glucosidase activities of regulatory mutants and their hybrids a Strain

L13 L19 L13-1 L19-2 L13-6 T3 L13-46 L19-6 L13-48 T1 L13 x L19 L13-1 x L19 L13-6 x L19 L13-46 x L19 T1 x L13 L13-48 x L19 L13-1 x L19-2 L13-1 x L19-8 c L13-17 x L19-14 c L13-5 x T3 L13-46 x L19-6 L13-48 x L19-2 L13-6 x L19-2 L13-46 x L19-2 L13-46 x L19-13 d L13-46 x T3 T1 x L13-1 T1 x L13-6 T1 x L13-46 T1 x Lt3-48

Genotype

Initial Initial rfp80 rfp80 rfp81 rfp81 RFP82 e RFP82 e

?(Semidominant) rfp82

Initial x Initial rfp80 x RFP80 rfp81 x RFP81 RFP82 e x RFP82 rfp82 x RFP82

Semidominant x Initial rfp80 x rfp80 rfp80 x rfp80 rfp80 x rfp80 rfp81 x rfp81 RFP82 c x RFP82 c Semidominant x RFP82 c rfp81 x rfp80 RFP82 c x rfp80 RFP82 c x rfp80 RFP82 c x rfp81 rfp82 x rfp80 rfp82 x rfp81 rfp82 x RFP82 e rfp82 x Semidominant

RF permease

c~-Glucosidase

Induced

Uninduced

Induced

Uninduced

24.8 24.8 29.6 23.0 22.5 26.6 30.1 21.3 25.3 0.7 36.7 36.6 39.0 31.9 24.8 27.2 22.5 34.9 29,9 22.5 24.8 23.6 22,5 28.4 23.0 37.2 17.7 22,2 34.3 30.1

0.7 1.3 29.6 21.3 16.0 21.3 35.5 23.0 26.0 1.0 1.7 1.8 1.8 24.8 1.9 3.9 17.1 2.9 5.8 11.2 23.0 28.4 1,4 27.7 8.6 40.7 1.8 2.9 40.8 4.2

31.4 ND b 34.7 ND 28.8 ND 32.6 ND 29.8 7.5 31.5 32.6 28.2 30.1 ND 33.4 28.9 31,5 ND ND ND 29.2 29.8 ND 28.8 ND ND ND ND ND

1.2 ND 32,4 ND 28.2 ND 29.0 ND 29.3 8.2 2.6 3.8 1.8 29.0 ND 6.6 32.4 17,4 ND ND ND 28.6 3,1 ND 19.9 ND ND ND ND ND

a

Cells were cultivated in liquid medium with 20 mg sucrose/ml (induced) or with 20 mg glucose/ml (uninduced) during 36 h at 30 °C with shaking. Concentration of RF in the medium with sucrose was 0.5 ~g/ml; in the medium with glucose - 0.5 ~g/ml (constitutive strains) or 50 gg/ml (inducible strains). Specific activities of RF permease and a-glucosidase were determined at least three times and were expressed as mean values in: ~moles of RF • 10 - 5 taken up by 1 mg of cells (dry weight) per 1 rain (RF permease) and ~moles of p-nitrophenol - 10 - 3 synthesized by 1 mg of permeabilized cells per 1 min (c~-glucosidase). Genotypes for markers other than indicated are omitted from the Table b Not determined c Interallelic complementation d Complementation

N e v e r t h e l e s s t h e segregation analysis c o n f i r m e d t h e d a t a o n allelism o b t a i n e d b y m e a n s o f t h e c o m p l e m e n t a t i o n test. In t h e analysis o f t h r e e h y b r i d s o b t a i n e d b y m e a n s o f crossing o f r f p 8 0 m u t a n t s b e t w e e n t h e m s e l v e s (L13-1 x L 1 9 - 2 ; L13-1 x L 1 9 - 8 ; L 1 3 - 1 7 x L 1 9 - 1 4 ; t h e l a t t e r t w o h y b r i d s are c h a r a c t e r i z e d b y interallelic c o m p l e m e n t a t i o n in R F P S O locus, see b e l o w ) o n l y c o n s t i t u t i v e segregants were observed, T h e y were also the only ones found among the meiotic progeny of the hybrid o b t a i n e d b y crossing o f t w o r f p 8 1 m u t a n t s : L13-5 x T3 ( T a b l e 2). A t t h e same t i m e diploid r f p 8 1 x r f p 8 0 ( L 1 3 - 6 x L 1 9 - 2 ) segregated a small n u m b e r o f c o n s t i t u t i v e strains (Table 2); it suggests t h e a b s e n c e o f linkage o f t h e r e l e v a n t loci.

The study of dominant and semidominant constitutive m u t a n t s was carried o u t b y t h e segregation analysis. O f 2 , 2 2 5 m e i o t i c segregants o f t h e h y b r i d o b t a i n e d b y crossing o f d o m i n a n t c o n s t i t u t i v e m u t a n t s ( L 1 3 - 4 6 x L19-6), 2 , 2 2 3 ones possessed c o n s t i t u t i v e R F p e r m e a s e ( T a b l e 2); t h u s suggesting close linkage, a p p a r e n t l y an allelism o f r e g u l a t o r y m u t a t i o n s in t h e strains tested. H e n c e t h e gene o f positive t y p e o f a c t i o n h a s b e e n f o u n d . T h e d o m i n a n t m u t a t i o n s in t h i s gene r e s u l t e d in t h e constitutive s y n t h e s i s o f R F p e r m e a s e ( t h e gene was desi g n a t e d as R F P 8 2 ; t h e m u t a t i o n s as R F P 8 2 e ) . T w o o u t o f 2 , 2 2 5 segregants o f h y b r i d L 1 3 - 4 6 x L 1 9 - 6 ( R F P 8 2 C x R F P 8 2 c) were u n a b l e t o s y n t h e s i z e R F p e r m e a s e even in t h e p r e s e n c e o f an i n d u c e r , sucrose.

111

A. A. Sibirny and G. M, Shavlovsky: Riboflavin permease and c~-glucosidaseregulation in yeast Table 2. Random spore analysis of hybrids obtained by crossing of various regulatory mutants for RF permease synthesisa Hybrid

Genotype

Total number of segregants tested

Number of constitutive segregants

L13-1 x L19 L13-6 x L19 L13-46 x L19 L13 x L19-6 L13-6 x L19-2 L13 x T1 L13-1 x L19-2 L13-1 x L19-8b L13-17 x L19-14b L13-5 x T3 L13-46 x L19-6 L13-46 x T1 L13-48 x L 1 9 - 6 L13-46 x L19-13

rfp80 x RFP80 rfp81 x RFP81 RFP82 c x RFP82 RFP82 x RFP82c rfp81 x rfpSO RFP82 x rfp82 rfp80 x rfp80 rfp80 x rfpSO rfp80 x rfpSO rfp81 x rfp81 RFP82c x RFP82 c RFP82c x rfp82 Semidominant x RFP82c RFP82 c x rfp80

599 1,022 550 232 730 975 1,553 4,883 5,210 2,894 2,225 1,027 1,488 459

220 117 291 144 288 1,553 4,883 5,210 2,894 2,223 873 1,343 375

%

36.7 11.4 52.9 62.1 39.5 100 100 100 100 99.9 85.0 90.3 81.7

Number of uninducible segregants

%

422 2 154 -

43.3 0.09 15.0 -

a For detail see: "Materials and Methods". Genotypes for markers other than indicated are omitted from the Table and were not analyzed b Interallelic complementation

The properties of one of them designated as T1 (mat +, rib2, argX) were studied in detail. The inability to synthesize RF permease by segregant T1 has been found to be a recessive marker since hybrid T1 x L19 produced RF permease inducibly and the hybrid of strain T1 with mutant L13-46 (RFP82 c) synthesized this transport system constitutively (Table 1). The analysis of meiotic segregation of the latter diploid showed the allelism of uninducibility (strain T1) and constitutivity (strain L13-46): among 1,027 segregants analyzed no one possessed a recombinant (inducible) phenotype (Table 2). Apparently uninducibility of RF permease in segregant T1 is a result of interallelic recombination in regulatory locus RFP82 among alleles RFP82C-1 (strain L13-46) and RFP82C-2 (strain L19-6). The genotype of segregant T1 was designated as rfp82. To study the properties of semidominant constitutive m u t a n t L13-48, it was crossed with m u t a n t L19-6 (RFP82 e) and random spore progeny of resulted hybrid was analyzed. Of 1,488 segregants tested, 145 (9.7%) were inducible and the rest 1,343 (90.3%) constitutive ones (Table 2). No uninducible segregants (analogous to segregant T1) have been found. These data suggest the absence of allelism of mutation RFP82 c and mutation in strain L13-48. Thus, complementation and segregation analyses of constitutive for RF permease synthesis mutants revealed at least three regulatory loci, two of which (RFP80, RFP81) act negatively and one (RFP82) positively. Qualitative testing (by colour of colonies) of phenotypes of constitutive mutants and their hybrids is very simple procedure. Hence we decided to carry out a corn-

plementation analysis of a greater number of constitutive mutants. 581 constitutive mutants selected from strain L13 were used. Each mutants was crossed with inducible strain L19 and with recessive mutants L19-2 (rfp80) and T3 (rfp81), and phenotypes of resulted hybrids were analysed. As shown in Table 3, 27 constitutive mutants were dominant, 21 were semidominant. The rest of constitutive mutants were recessive. The complementation analysis showed that 370 of them bore mutations in locus RFP80 and 163 ones in locus RFP81. We did not observe any recessive constitutive m u t a n t belonging to another complementation class. The number of genes, mutations in which are dominant or semidominant and result in the constitutive RF permease synthesis were not determined.

Mutations with cold-sensitive products o f regulatory genes. Interallelic and intergenic complementation o f constitutive mutants To determine the nature of regulatory loci RFP80, RFP81 and RFP82, we attempted to reveal mutations in the loci with a temperature-sensitive phenotype. It was shown that of 74 constitutive mutants, isolated from strain L13 and used for the complementation and segregation analyses, 11 mutants rfp80 and 1 mutant rfp81 synthesized RF permease constitutively at 27 °C but not at 35 °C. At elevated temperature these mutants formed RF permease only in the presence of the inducer (sucrose). The results obtained suggest a cold-sensitivity

112

A.A. Sibirny and G. M. Shavlovsky: Riboflavin permease and c~-glucosidaseregulation in yeast

Table 3. Genotypes of constitutive mutants for RF permease synthesis selected from strain L13

Table 4. The allelic specificity of the restoration of inducible RF permease synthesis in diploids obtained by crossing of constitutive rfp80 and RFP82e mutants a

Manifestation in heterozygote

Genotype

rfp80 mutant

Dominant Semido minant Recessive Recessive

RFP82e (?) ? rfp80 rfp81

Number of mutants

%

27 21 370 164

4.6 3.6 63.6 28.2

o f the products of RFPSO and RFP81 loci in these strains and indirectly point to their protein nature. Complementation analysis revealed the restoration of a partly wild (inducible) phenotype in some rfp80 x rfp80 diploids (Table 1). This also suggests the protein nature o f the product o f regulatory locus RFPSO. Interallelic complementation was not observed in rfp81 x rfp81 diploids; however it should be pointed out that all the rfp81 mutants isolated from strain L13 (181 mutants) were crossed with one rfp81 mutant o f the opposite mating type, i.e. T3 (segregant o f diploid L13-6 x L19). Using such limited material we did not observe interallelic complementation in RFP81 locus. Identification o f dominant RFP82 e mutants was based on the determination o f constitutive hybrids formed b y crossing these mutants with initial inducible strains. Unexpectedely it has been found that some hybrids obtained b y crossing o f both RFP82 c mutants (L13-46 and L19-6) with some constitutive rfp80 mutants possess a partly restored ability for inducible R F permease synthesis (Table 1: compare hybrids L13-46 x L19-2 and L13-46 x L19-13). The mutual allelic specificity o f inducible phenotype restoration was observed b y interaction o f RFPSO and RFP82 genes (Table 4). This is a genetic evidence for interaction o f corresponding gene products (Inge-Vechtomov and Soidla 1978). The restoration o f inducible phenotype in some RFP82 c x rfp80 hybrids suggests a complementation o f mutually damaged products o f regulatory loci RFPSO and RFP82 o f the negative and positive types of action, respectively. This can be easy explained when the products o f indicated loci appear to be proteins.

Regulatory gene R F P 8 2 is epistatic to gene RFPSO To answer the question about the succession o f actions o f identified regulatory genes, uninducible segregant T1 (rfp82) was crossed with constitutive mutant L13-1 (rfp80), and the meiotic segregation o f the obtained hybrid was analyzed. Segregants unable to induce R F permease were crossed with rfp80 mutants o f both

RFP82e mutant L13-46

T8

Hybrid phenotype L19-4 L19-13 L19-16 L19-17 L19-24 L19-25 L19-27 L19-29 L19-33 L19-34 L19-35 L19-36 L19-37 L19-38 L19-39 L19-40 L19-42 L19-43 L19-44 L19-45 L19-46 a

y

w

w w

y y

w

w

w

w

y y y w w w w w w w

w w w y y y y y y y

w

w

w w w w w

y y y y y

46 rfp80 mutants isolated from strain L19 were crossed with two RFP82e mutants of opposite mating type: L13-46 and T8 (RFP82e segregant of hybrid L19-6 x L13). This Table includes the data on hybrid phenotype of 24 rfp80 mutants; the rest rfp80 mutants tested (25 strains) produced the constitutive hybrids after crossing with both RFP82e mutants. "y" - yellow colour of hybrid colonies after three days of growth on the medium with glucose (5 g/l) and RF (200 ~zg/ml) (constitutive synthesis of RF permease). "w" - white colour of hybrid colonies on the same medium (restoration of inducible RF permease synthesis)

mating types. Among 11 segregants tested, there were two which though were uninducible, yet contained mutation rfp80 since the isolated hybrids constitutively produced RF permease. Thus gene RFP82 is epistatic to gene RFPSO; i.e. RFP82 gene product (positive-acting) acts after RFP80 gene product (negative-acting). An analogous experiment was carried out with uninducible segregants o f hybrid T1 x L13-6 (rfp82 x rfp81) which were crossed with constitutive rfp81 mutants o f b o t h mating types. The phenotypes o f resulted hybrids were analysed. Of 26 uninducible segregants there was no one able to form constitutive diploids after being crossed with T3 or L13-6 (rfpS1) mutants. Therefore the succession o f RFPS1 and RFP82 gene products is still an unanswered question.

A. A. Sibirny and G. M. Shavlovsky:Riboflavin permease and a-glucosidase regulation in yeast The same loci regulate synthesis o f R F permease and a-glucosidase It is known that RF permease of P. guilliermondii is formed coordinatively with the enzyme of a-glucosides utilization, i.e. a-glucosidase (Sibirny et al. 1979). It was interesting to elucidate the pattern of this enzyme synthesis in different classes of constitutive mutants for RF permease formation. Table 1 shows that all classes of constitutive for RF permease synthesis mutants possess also a constitutive character of the a-glucosidase synthesis (rfp80, rfp81, RFP82 c, semidominant mutant L13-48). The hybrids tested possessed also a similar patterns of RF permease and a-glucosidase syntheses. InteralMic complementation in RFPSO locus as well as complementation of some RFP82 e x rfp80 hybrids result in the inducibility of the synthesis not only of RF permease but also that of a-glucosidase (Table 1). All these data clearly indicate to the participation of RFP80, RFP81 and RFP82 loci in regulation of c~-glucosidase synthesis. The results of determination of c~-glucosidase activity in uninducible for RF permease synthesis segregant T1 (rfp82) are of a special interest. It was shown (Table 1) that this strain possesses low level of a-glucosidase activity (though sufficiently increased as compared with non-induced cells of initial strains); and this enzyme is synthesized constitutively. These data suggest that: (i) a lesser concentration of positive-acting RFP82 product is sufficient for a cell to synthesize a-glucosidase rather than RF permease, or (ii) the targets for RFP82 gene product action (operator loci of RF permease and a-glucosidase structural genes?) differ from each other in the structure. It is worth mentioning that in hybrids rfp80 x rfp80 and rfp80 x RFP82 c, characterized by inducible RF permease synthesis, the ratios of a-glucosidase levels in the inducer-free medium, compared to the levels in the medium with the inducer, are considerably higher than the corresponding ratios for RF permease (Table 1).

Discussion

The results presented indicate the existence of the multicomponent cascade system of regulatory genes involved in expression of RF permease and a-glucosidase in the yeast P. guilliermondii. Similar cascade systems were identified earlier in the yeast Saccharomyces cerevisiae participating in genetic control of the formation of acid phosphatase (Kozhin a. Ter-Avanesian 1979; Toh-e et al. 1981) as well as enzymes of degradation of galactose (Perlman and Hopper 1979), allantoin (Cooper 1982) and arginine (Wiame 1973; Dubois et al, 1978). For each such system the regulatory genes of positive and negative

113

types of action were identified and severe hierarchy of their interaction was found. We identified three regulatory genes of P. guiIliermondii involved in the synthesis of RF permease and aglucosidase, two of which (RFP80, RFP81) acted negatively and one (RFP82) positively. From the data on the detection of temperature-sensitive regulatory mutants as well as on interallelic complementation of some rfp80 mutants, one may assume that the products of RFP80 and RFP81 genes are proteins. Epistasis-hypostasis test showed that negativelyacting gene RFPSO acts prior to positively-acting gene RFP82. How does gene RFPSO switch off the action of gene RFP827 Based on the data on the mutual allelie specificity of regulatory genes of positive and negative control (RFP82 and RFP80, respectively) it is easy to suggest a direct protein-protein interaction of products of these regulatory genes. In inducible strains in the absence of inducer, RFP80 product inactivates RFP82 product which is absolutely necessary for the expression of corresponding structural genes; in the presence of inducer, RFP80 product is inactivated and cannot bind RFP82 product. In rfp80 mutants, the affected product of this gene does not inactivate RFP82 product independently of the presence of inducer. RFP82 e mutants contain the impaired positively-acting product which does not interact with RFPSO regulatory protein; this results in the constitutive expression of RF permease and a-glucosidase. In rfp82 strains, the affected RFP82 gene product either binds even with inactive rfpSO product or cannot effectively promote the expression of the corresponding structural genes. In some RFP82 c mutants, the affected regulatory protein not interacting with wild RFP80 product can be bind, in the absence of inducer, with the mutationally impaired product of some rfp80 alleles, this results in inactivation of positively-acting regulatory protein and restoration of inducibility. The model of interaction of regulatory proteins involved in the expression of RF permease and a-glucosidase is presented in Figure. Formally it resembles the models of interaction of S. cerevisiae gene products participating the regulation of syntheses of acid phosphatase (Toh-e et al. 1981; Oshima 1982) and enzymes of galactose degradation (Perlman and Hopper 1979; Matsumoto et al. 1980). For each system the mutations in the positively-acting regulatory gene were found which resulted in the constitutive synthesis and uninducibility of corresponding products of structural genes (RFP82 c - rfp82; PH082 - pho4; GAL81 -gal4). Unfortunately in our model there is no place for gene RFP81. To determine the role of this gene further studies are necessary. The possibility to identification of new positively-acting regulatory genes also exists. To confirm the above model of participation of regulatory genes in RF permease and e-glucosidase syntheses,

114

A.A. Sibirny and G. M. Shavlovsky: Riboflavin permease and c~-glucosidaseregulation in yeast

the isolation o f dominant uninducible mutations in

RFPSO and RFP81 loci as well as mutations in structural loci is desirable. The nature of described RF permease and the reasons causing coordinate regulation of the syntheses of this transport system and a-glucosidase are unknown. One may assume that in wild strains ofP. guilliermondii, the structural gene of RF permease is not expressed because of the absence of a suitable promoter. Under certain conditions of selection (isolation of RF deficient strains growing at very low RF concentrations) the transfer of the silent structural gene to rather a strong promoter may occur (to promoter of a-glucosidase or a-glucoside permease in this case). Such "gene engineering in vivo" resulted in the dependence of RF permease synthesis on a-glucosides.

References Cooper TG (1982) In: Hollander A (ed) Genetic engineering of microorganisms for chemicals. Plenum Press, New York, pp 143-161 Dubois E, Hiernaux D, Grenson M, Wiame J-M (1978) J Mol Biol 122:383-406 Inge-Vechtomov SG, Soidla TR (1978) In: Glotov NV, Poroshenko GG (eds) General genetics, vol 3. Evolutional and populational genetics (in Russian). ltogi nauki i tekhniki. VINITI AN SSSR, Moscow, pp 7-37 Kozhin SA, Ter-Avanesian MD (1979) Issledovaniya po genetike (USSR) No. 8, pp 89-109 Matsumoto K, Adachi Y, Toh-e A, Oshima Y (1980) J Bacteriol 141:508-527

Oshima Y (1982) In: Strathern JN, Jones EW, Broach JR (eds) The molecular biology of the yeast Saeeharomyees. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, pp 159-180 Perlman D, Hopper JE (1979) Cell 16:89 95 Shavlovsky GM, Sibirny AA, Kshanovskaya BV, Koltun LV, Logvinenko EM (1979) Genetika (USSR) 15:1561-1568 Shavlovsky GM, Sibirny AA, Ksheminskaya GP (1977) Biochem Physiol Pflanzen 171 :139-145 Sib[rny AA, Ksheminskaya GP, Orlovskaya AG, Shavlovsky GM (1981) Biokhimiya (USSR) 46:1761-1763 Sibirny AA, ShavlovskyGM, Goloshchapova GV (1977a) Genetika (USSR) 13:872-879 Sibirny AA, Shavlovsky GM, Kshanovskaya BV, Naumov G1 (1977b) Genetika (USSR) 13:314-321 Sibirny AA, Shavlovsky GM, Ksheminskaya GP, Orlovskaya AG (1977c) Mikrobiologiya (USSR) 46:376-378 Sibirny AA, Shavlovsky GM, Ksheminskaya GP, Orlovskaya AG (1977d) Biokhimiya (USSR) 42:1841-1851 Sibirny AA, Shavlovsky GM, Ksheminskaya GP, Orlovskaya AG (1978) Biokhimiya (USSR) 43:1414-1422 Sibirny AA, Shavlovsky GM, Ksheminskaya GP, Orlovskaya AG (1979) Biokhimiya (USSR) 44:1558-1568 Sibtrny AA, Zharova VP, Kshanovskaya BV, Shavlovsky GM (1977e) Tsitologiya i genetika (USSR) 11:330-333 Toh-e A, Inoue S, Oshima Y (1981) J Bacteriol 145:221-232 Wiame J-M (1973) In: Suomalainen H, Waller C (eds) Proc 3d Intern Spec Symp on Yeasts, part If. Print Oy, OtaniemiHelsinki, pp 307-330

Communicated by S. G. Inge-Vechtomov Received July 7 / October 6, 1983