Eur. J. Biochem. 194,293-299 (1990) 0FEBS 1990

Photoaffinity labelling of the purine - cytosine permease of Saccharomyces cerevisiae Maria-Chantal CHIRIO, Daniel BRETHES, Christian NAPIAS, Xavier GRANDIER-VAZEILLE, Felicien RAKOTOMANANA and Jean CHEVALLIER Institut de Biochimie Cellulaire et de Neurochimie du Centre National de la Recherche Scientifique, Bordeaux, France (Received June 6/August 22, 1990) - EJB 90 0643

8-Azidoadenine was used as a photoaffinity reagent to characterize the purine - cytosine permease of Saccharomyces cerevisiae. It is a potent competitive inhibitor of cytosine uptake and irradiation of the cells incubated with the label induced the irreversible inactivation of cytosine uptake. Addition of excess cytosine prevented this labelling which was restricted to the outer face of the plasma membrane since it was not accumulated by the cells. In the strain with the amplified purine-cytosine permease gene the maximum cytosine uptake rate was increased 4 - 5-fold relative to wild type without a modification of the Michaelis constant of uptake ( K J ; no uptake could be measured in the deleted strain. The relative amounts of specific labelling determined for the cells and for membrane preparations were 0, 1 and 4 for the null, the wild-type and the amplified strains, respectively. One major band specifically labelled by [3H]azidoadenine, corresponding to a polypeptide with an apparent molecular mass of 45 kDa, was observed in the wild type, amplified in the strain carrying the multicopy plasmid and not detected in the deleted strain. Therefore this polypeptide corresponds to the purine -cytosine permease.

In the yeast Saccharomyces cerevisiae, at least two transport systems are involved in the uptake of purine and pyrimidine bases: one for uracil [l] and the other for purine bases (adenine, guanine and hypoxanthine) and a pyrimidine base (cytosine) [2 -41. This purine -cytosine permease is believed to work as a proton symporter and cation antiporter through the dissipation of an electrochemical gradient of protons built up by the H+-ATPase located in the membrane [5, 61. Study of this permease mechanism is now facilitated by recent data on the nucleotide sequence of the gene, the FCY2 gene [7], and by the opportunity to increase the amount of this protein in the yeast cell using a cloned permease gene on a multicopy plasmid [7, 81. As a first step in the study of the purine-cytosine permease mechanism, we describe conditions allowing covalent specific photoaffinity labelling of a protein involved in cytosine uptake by using 8-azidoadenine, a derivative of one of the transporter ligands [9]. Labelling was performed either in vivo (on cells) or in vitro (on plasma membrane) on different S. cerevisiae strains. After kinetic analysis of the action of the photoaffinity label on cytosine transport, we have studied the correlation between the specific covalent labelling of the cells and the inactivation of transport activity. We then measured the amount of polypeptide specifically labelled and involved in the uptake directly on cells and on membrane preparations Correspondence to J . Chevallier, Institut de Biochimie Cellulaire et de Neurochimie du Centre National de la Recherche Scientifique, 1 rue Camille Saint-Saens, F-33077 Bordeaux Cedex, France Abbreviations. N,Ade, 8-azidoadenine; N3[,H]Ade, 8-azido[23H]adenine; PhMeS02F, phenylmethylsulfonyl fluoride; YNB wjo, yeast nitrogen base without amino acids. Enzymes. Mitochondria1 ATPase, ATP phosphohydrolase (EC 3.6.1.34); H+-ATPase adenosine triphosphatase (EC 3.6.1.35); cytosine deaminase (EC 3.5.4.1).

isolated from the different strains. In addition, we have provided evidence for one main polypeptide chain with an apparent molecular mass of 45 kDa in the same range as the molecular mass of the permease calculated from the nucleotide composition of the gene [7]. This polypeptide is probably the purine -cytosine permease encoded by the FCY2 gene. EXPERIMENTAL PROCEDURES Materials

Adenine, adenosine and cytosine were purchased from Aldrich and 8-azidoadenosine from Sigma. The azidoadenine derivative was obtained by acid hydrolysis of 8-azidoadenosine [lo]. [2-3H]Adenosine (850 - 920 GBq/mmol) was obtained from Amersham and [2-14C]cytosine(1.6 - 2.2 GBq/ mmol) from the Commissariat 1’Energie Atomique. 8Azid0[2-~H]adeninewas either synthesized in the laboratory (see below), with a specific radioactivity at 5.6 GBq/mmol, or obtained after acid hydrolysis of 8-azid0[2-~H]adenosine (740 - 850 GBq/mmol) from Moravek Inc. Yeast nitrogen base without amino acids (YNB w/o) was from Difco. Prestained SDS molecular mass markers were obtained from Sigma. Scintillation cocktail was Ready Value from Beckman and fluorography enhancer was Amplifier from Amersham. All other reagents were of analytical grade from Merck. Strains and media

FL442-4B mat a fcyl-1 (cytosine-deaminase-less) was isolated by Jund and Lacroute [ll]. NC233-10B leu2 fcyl-1 fcy2A(BglII -KpnI) was constructed by crossing a strain having a deletion of 0.65 kb in the open reading frame of the purine -cytosine permease chromosomal gene [7] with strain NC219-4B leu2,fcyl-1. The resulting strain NC233-10B was

294 then transformed either with plasmid pJDB207 a S. cerevisiael E. coli shuttle vector which consists of pBR322, and the following S. cerevisiae sequences: the LEU2 gene and the EcoRID segment of the 2-pm plasmid carrying the DNA replication origin [12] or with plasmid pAB which derives from pJBD207: a ClaI - SphI DNA segment of 2.4 kb from S. cerevisiue containing the purine - cytosine permease gene (FCY2) was inserted at the corresponding sites of the pBR322 part of the vector [8]. Strain NC233-10B(pJDB207) remains a permeasenull strain whereas NC233-10B(pAB) is a permease-proficient strain carrying multiple copies of the FCY2 gene. In the text, the NC233-10B transformed clones are referred to only by the plasmid they harbor, i.e. pJDB207 or pAB, and FL442-4B was the wild type. All these strains are cytosine-deaminaseless @yI-I). This mutation prevents metabolic utilization of cytosine and allows correct determination of uptake rate. Strains were grown at 30' C under agitation in a YNB w/o medium (6.75 g/l) containing 2% glucose and 25 mM sodium phthalate pH 5.5. The cells were harvested during the exponential phase (5 -- 7 x 1O7 cells/ml).

Plasma membrane preparution

Plasma membranes were prepared as described in [15] with some modifications. Wet cells (40 g) harvested during the exponential phase were resuspended in 40 ml 50 mM imidazole pH 7.5 containing 1 mM MgCl,, 0.25 M sucrose and protease inhibitors (0.2 mM PhMeSO'F, 5 mM p-aminobenzamidine and 2 pg/ml of chymostatin). The suspension was ground at 4°C by vortexing in the presence of 400g 0.45-mm glass beads for 210 s at top speed in a Frisch grinder [16]. The resulting material was fractionated by centrifugation at 4 "C. The supernatants were successively centrifuged twice at 2500 g for 5 min, once at 7500 g for 5 min and once at 15000 g for 45 min. The resulting pellet was resuspended at about 3 mg protein/ml in 40 ml 10 mM imidazole pH 7.5 containing 0.25 M sucrose with 0.2 mM PhMeSO'F, 5 mM paminobenzamidine and 0.2 pg/ml of chymostatin. Mitochondrial material was eliminated by acid precipitation [15]. The pH of the suspension was quickly lowered to 4.7 by addition of concentrated acetic acid. This suspension was immediately centrifuged at 7500 g for 45 s. The pellet was discarded and the supernatant was rapidly adjusted to pH 7.5 with concentrated Synthesis of8-uzid0(2-~H]adenine NaOH and centrifuged at 140000 g for 40 min. The resulting 8-Azid0[2-~H]adeninewas synthesized in three steps from pellet (30 - 40 mg protein), which was resuspended at 5 [2-3H]adenosine (62 pmol, 5.6 GBq/mmol) as starting ma- 6 mg/ml in 10 mM imidazole pH 7.5 containing 5 mM p-amiterial. 8-Brom0[2-~H]adenosine,the intermediary product of nobenzamidine and 2 pg/ml chymostatin, represented the the process, was synthesized in aqueous medium as described crude plasma-membrane-enriched fraction used throughout [13] and purified by elution through a DEAE-cellulose column this work. A similar procedure adapted to smaller quantities (25 x 1 cm). Fractions were analysed by thin-layer chromatog- of cells was used when membranes were prepared from cells raphy on precoated plates (silica gel 60 F-254 from Merck) previously labelled by N3[3H]Ade. In this case, 6 x lo9 cells using 1-butanol/acetic acid/water (77: 13: 10) as eluant. The (0.25 g wet mass) in 200 ml were labelled with 15 pM fractions containing only 8-bromoadenosine were selected and N3[3H]Ade prior to membrane preparation yielding about pooled. Reaction yield was of the order of 31%. Solvent was 300 pg labelled membrane fraction. The ATPase activity of the preparations was routinely removed under vacuum and the product was dried on P 2 0 5 , then dissolved in 1 ml anhydrous dimethylformamide and 1.2 pmol . min-' . mg-I protein at pH 5.8 and 30°C. This kept in the dark at 80°C with 200 pmol NaN3 under agitation activity was inhibited at 90% by adding 100 pM vanadate. [9]. Under these conditions, the reaction was completed within The amount of contamination from mitochondria1 material was determined by oligomycin inhibition of the oligomycin16 h. The resulting 8-azid0[2-~H]adenosine 38000 M - ' cm-' in 0.1 M HCI) wasconcentrated to dryness, sensitive ATPase at pH 8.5 measured immediately after the then dissolved in 0.5 ml 0.5 M HCl and finally hydrolysed to preparation. This contamination never exceeded 10%. The 8-azid0[2-~H]adenineby heating at 100°C for 3 h in the dark ratio between vanadate-sensitive ATPase at pH 5.8 and [lo]. Under these conditions, hydrolysis was completed, and oligomycin-sensitive ATPase at pH 8.5 was of the order of 3 the final solution contained almost pure 8-azid0[2-~H]adenine 4, values similar to those already obtained by other authors [15, 17, 181. ( ~ 2 8 3= 20000 M - ' cm-' in 0.1 M HCl). No trace of 8-bromoadenine (c264.5 = 17100 M - ' cm-' in 0.1 M HC1) was detected. Routinely the overall yield of synthesis was of the order of 23% and the specific radioactivity was 5.6 GBq/ Irrudiu t ion experiments mmol. Cells (3 x lo7 cells/ml) in 50 mM sodium citrate pH 5.5 containing 2% glucose and 100 mM NH4Cl were set at 10 cm Cytosine uptake from a Bioblock Scientific Mineral lamp (312 nm, 1000 pW/ For measurements of cytosine uptake, the cells were resus- cm') and irradiated under agitation for 2 rnin at 4°C in the pended at 5 - 7.5 x lo6 cells/ml in 50 mM sodium citrate presence of N3Ade or N3[3H]Ade at different concentrations. pH 5.5 containing 2% glucose at 30°C. The reaction was After irradiation, the cells were centrifuged at 10000 g for started by addition 0f[2-'~C]cytosine(110- 170 MBq/mmol). 5 rnin and washed twice with cold water. The pellets were As a function of time, aliquots (0.3 ml) were filtered through assayed for cell number, radioactivity and cytosine uptake. 0.8-pm Millipore filters, washed twice with 3 ml cold water, Membrane preparations (1 mg/ml) were irradiated under the dried and assayed for radioactivity measurements. Total same conditions and collected by centrifugation at 120000 g cytosine concentrations in these assays varied over 0.2 - for 15 rnin in a Beckman TLA 100.2 rotor. The pellets were 100pM. Uptake rates were determined either by using a resuspended in the sodium citrate buffer and assayed for proquenched-flow apparatus [14] for times shorter than 8 s or tein content and radioactivity. Non-specific binding was determanually for longer times. For the study of cytosine uptake mined on cells or membrane preparations under the same inhibition by azidoadenine, the assays were performed as conditions but in the presence of 2 mM cytosine. NH4CI was above except thiit they were done in the dark in the presence added to the irradiation medium in order to minimize the nonof N3Ade (0.5- 100 pM) in the assay medium. specific binding.

29 5 SDS/polyacrylamide gel electrophoresis

Slabs of gel (15 x 10 x 0.75 cm) containing a polyacrylamide gradient ( 5 - 20% mass/vol.) were made according to Laemmli [19]. Before loading onto the gels, the samples (2 mg/ ml of protein) were resuspended in 300 mM Tris/HCl pH 8.9 containing 5% (massivol.) SDS, 20% (by vol.) glycerol, 0.05% (massivol.) bromophenol blue and 5% (by vol.) 2-mercaptoethanol and incubated for 15 min at 37°C. Immediately after the run, the gels were cut between wells in lanes, which were then sliced. Each slice (0.5 cm) was treated overnight with 0.5 ml 30% (by vol.) H 2 0 2at 60°C. Radioactivity was measured by scintillation counting in 6 ml liquid scintillation cocktail containing 0.7% acetic acid. Counting efficiency was usually about 40%, as determined from quenching calibration curves. For fluorographic detection, gels were soaked for 30 min in Amplifier, then dried and exposed to Amersham hyperfilm HP at -80°C for 10 or 21 days.

a,

c

1

'5 &I

0

60

90

120

150 180

time (s)

Miscellaneous

Cell concentration was determined by measuring turbidity at 550 nm, an absorbance of 1 corresponding to 3 x lo7 cells/ ml. Protein concentration was determined by the method of Lowry [20] with bovine serum albumin as standard. ATPase activity was measured at 30°C at initial velocity in two different conditions: (a) oligomycin-sensitive ATPase in 30 mM Tris/HCl pH 8.5 containing 5 mM ATP, 5 mM MgC12, 100 mM KCl, with or without 2.5 pg/ml oligomycin (for the determination of contamination by mitochondria1 material); (b) H+-ATPase in 50 mM Mes/Tris pH 5.8, 5 mM ATP, 10 mM MgC12, with or without 100 pM vanadate (for plasma membrane H +-ATPase measurements). The reactions were started by adding ATP and stopped by adding trichloroacetic acid to 5% final concentration. The inorganic phosphate concentration was determined as described by Ottolenghi [21]. Radioactivity measurements were performed on a Beckman LS 3801 scintillation counter.

30

\ 3 2 i

. s 1

' 0

0

1 2 3 V/[SI (ml . min-1. I 07ce11s-1)

Fig. 1. Cytosine uptake by different strains. Cells from different cytosine-deaminase-less strains, i.e. (0)FL442-4B (wild type), ( A ) pJDB207 (permease-null) and ( W , 0)pAB (carrying the cloned purine - cytosine permease gene FCY2) harvested in the exponential phase, were washed in 50 mM sodium citrate pH 5.5 containing 2% glucose at 30°C and assayed for cytosine transport. (A) Uptake was started by addition of [2-'4C]cytosine (110 MBq/mmol) at a final concentration of 50 pM. At different times, aliquots of 0.3 ml were filtered as described under Experimental Procedures. For pAB strain, the initial phase of uptake (between 2-10 s) was determined in the same conditions using a quenched-flow apparatus (0).(B) EadieHofstee plots of cytosine uptake of pAB ( W ) and FL442-4B (0) strains. Initial velocities of uptake were determined for cytosine concentrations varying over 0.2 - 140 pM

RESULTS Kinetic studies h

The kinetics of cytosine uptake by strains pAB, FL442-4B and pJDB207 are presented in Fig. 1. As expected, cells from the permease-null strain do not show any transport activity and the rate of cytosine uptake is higher for the strain in which the cloned permease gene was introduced than for the wild type. As shown by quenched-flow measurements, the kinetics of cytosine uptake are constant up to 10 s of reaction time under our experimental conditions (Fig. 1A). Therefore, in this paper, initial rates were measured after 8 s incubation. The maximal rate of cytosine uptake (Vm) determined from Eadie-Hofstee plots (Fig. 1 B) was 5 times greater in the strain with the amplified FCY2 gene than in the wild-type strain: 6.0 & 0.3 nmol . min-' . lo7 cells-' and 1.2 f 0.1 nmol . min-' . lo7 cells-' respectively. The apparent Michaelis constant of uptake (K,) was similar for the two strains (1.8 f 0.2 pM) (Fig. 1B). When incubated in the dark, under conditions where no covalent fixation could occur, N,Ade was a competitive inhibitor of uptake with a Ki of 3 pM (Fig. 2). When the cells were irradiated at 4°C in the presence of N,Ade, extensively washed, then assayed for cytosine uptake at 30"C, a decrease in maximal uptake rate was observed (Fig. 3). Control exper-

7

> 0

1

2

3

V/[SI (ml . min-1. 107ce11s-1)

Fig. 2.8-Azidoadenine as competitive inhibitor of cytosine uptake in the dark. Cytosine uptake was determined as described in Fig. 1 B at different concentrations of cytosine (0.2- 140 pM) in absence or in presence of N,Ade. Results are plotted according to Eadie-Hofstee. Concentrations of N,Ade: 0 (W), 6 pM (A),16 pM ( 0 )

iments showed that the irradiation products of the azido compound were not cytosine-uptake inhibitors; besides, the irradiation process had almost no effect on the cytosine uptake

296

-4

70-

pb

60-

8

0

1

V/[SI (mi . mi”-’.

2

3

0

“g[’HIAdel

IO~C~IIS-~)

Fig. 3. Irreversible inactivation of cytosine uptake after irradiation in presence qf N3Ade. Cells from pAB strain were resuspended to 3 x lo7 cells/ml in 50 mM sodium citrate pH 5.5 containing 2% glucose and 100 mM NH4CI arid incubated at 4°C in the absence ( W ) or in the presence ( 0 )of 36 pM N3Ade. Then samples were irradiated under agitation for 2 min. washed as described under Experimental Procedures, and assayed for cytosine transport as described in Fig. 1 . Results are plotted according to Eadie-Hofstee

40

20

60

(”4

Fig. 4. Quantification of specific labelling of cells by N3r3H/Ade. Irradiation of cells was performed as described in Fig. 3 with different concentrations of N3[3H]Ade (2-50 pM at 5.6 GBq/mmol) in the absence ( W , A) (total labelling) or in the presence (0, A ) of 2 mM cytosine (non-specific labelling). After irradiation, pAB ( W , 0)and pJDB207 (A,A ) cells were extensively washed and assayed for radioactivity as described under Experimental Procedures

(a S-10% decrease of V,,, was generally observed for the control). Since N,Ade was a competitive inhibitor of uptake and since the nitrene radical produced upon irradiation promoted an irreversible inactivation of uptake, it may be concluded that the inactivation is due to covalent linkage of this reagent at or near the active site of the permease. Therefore, N,Ade was used to analyse the relationship between the labelling and the inactivation of the transport activity, and to characterize the labelled polypeptide. Speci$c labelling oj’cells

At 3 0 T , N,[,H]Ade was slightly accumulated into the cells but 100-200 times less than cytosine. At 4”C, the temperature at which labelling experiments were performed, no uptake of the azido reagent could be detected. In conventional affinity labelling of cells, the usual way to distinguish between specific and non-specific labelling is to use a protecting agent, usually the natural ligand. A significant specific labelling was observed for pAB cells, as evaluated by the difference between the amount of radioactivity incorporated into cells irradiated, in the absence and in the presence of 2 mM cytosine (Fig. 4). For the permease-null strain pJDB207, no difference was observed between labelling measured in the absence or in the presence of 2 mM cytosine (Fig. 4). Therefore, we considered the amount of radioactivity bound onto the cells after irradiation in a buffer containing 2 mM cytosine as non-specific labelling. Non-specific labelling was linear as a function of N3[3H]Adeconcentrations (Fig. 4) and was quantitatively the same for the three strains. Other permease substrates, such as adenine or hypoxanthine, could be used instead of cytosine for the determination of non-specific labelling (data not shown). Since the irradiation conditions were optimized to preserve the uptake activity of the cells (Fig. 3, upper curve), and since it was possible to determine specific labelling using N3[3H]Ade (Fig. 4), we studied the relationship between inactivation and the amount of label specifically incorporated. Despite variations in the reproducibility of the irradiation process and cumulative experimental errors, a correlation between the decrease in V , and the quantity of specifically bound label was observed (Fig. S ) . Assuming that one molecule of N3Ade is

01

0

-.-

I

20

--. -. 40

60

80

N3[’H]Ade bound (prnolil 07cells)

Fig. 5. Relationship between decrease r,J’rnaximal rate o j cytosine uptake and specific labelling. Cells from FL442-4B ( 0 )and pAB ).( strains were incubated under the conditions described in Fig. 4. After irradiation, samples were washed and assayed as described under Experimental Procedures. Labelling determination and cytosine uptake measurement were done in triplicate on each sample. The amount of specific labelling was determined as the difference between total and non-specific labellings (see Fig. 4). Extrapolation for total inactivation (dotted line) gave a maximum number of specific binding sites for N3[3H]Ade of 90 pmol/lO’ cells for pAB and 30 pmol/lO’ cells for FL442-4B

reacting with one molecule of permease and inactivates it, a direct extrapolation of the data presented in Fig. 5 gives an indication of the maximal amount of active permease in cells: 30 pmol/107 cells and 90 pmol/107 cells for FL442-4B and pAB respectively. In addition the turnover number of uptake (which represented the number of cytosine molecules translocated/molecule permease within a unit of time) could be obtained from the slope of the straight lines (Fig. 5). It was the same for the two strains and of the order of 1 s-’. The characterization of the polypeptides specifically labelled by N3[3H]Adewas done on SDSjPAGE. The radioactivity profiles of gels, obtained with membrane samples isolated from cells previously labelled with N3[3H]Ade, are shown in Fig. 6. A large specific labelling of a polypeptide of an apparent molecular mass in the 45-kDa region was ob-

297

-E

2400 -

180 116 84 58 I

I

I

I

8 2000 -

I

1 1

48 36.5 266

v

A

I

91600-

8 ._

g n

1200:

1.41

h 1.2

1.04

2 800 -

5

400 0

-k ’D

1600

z 1200

0

5

10 15 20 slice number (5 mm)

25

180 116 84 58 48 36.526.6 I

I

1

I

1

I 1

0

c ._

>

.-

2

2 7 I

20

40 “3[3HlAdel

800 400

0

m

slice number (5 mm)

Fig. 6 . SDSjPAGE of proteins prepared from cells labelled by N 3 [ 3 H / A d e .Prior to membrane preparation, cells from pAB (A) and pJDB207 (B) were irradiated in the presence of 15 pM N,[3H]Ade (5.6 GBq/mmol) and in the absence ( 0 )or in the presence (0)of 2 mM cytosine; 50 pg plasma membrane fractions were submitted to SDS/PAGE. Each lane was sliced (5 mm width), treated as described and radioactivity measured. At the top of the figure are indicated the positions and the molecular mass values (in kDa) of prestained standards run on the same gel in a parallel lane. Specific labelling of pAB cells was 16 pmol/107 cells leading to 230 pmol/mg protein for the membrane fraction prepared from these cells

served in material isolated from the pAB strain (Fig. 6A). No specific labelling was detected for pJDB207 (Fig. 6B). Considering the paucity of plasma membrane in cells and the overall yield of membrane preparation, the amounts of cells and N3[3H]Adenecessary for labelling the cells were quite large. Therefore, further labelling experiments were performed directly on plasma membrane-enriched fractions isolated from the three strains. Specific labelling of rnernbrane,jractions

Labelling experiments were performed as described in Experimental Procedure. After correction for the non-specific labelling, the amount of specific labelling increased as a function of the concentration of N3[3H]Ade and reached a maximum value depending on the strain (Fig. 7). On average, these values were 200 100 pmol/mg protein for FL442-4B and 800 & 200 pmol/mg protein for pAB membranes, with no specific labelling detected for pJDB207 samples. These data indicate that the plasma-membrane-enriched fractions from wild-type and pAB strains still contain the protein capable of reacting specifically with the azido reagent. In addition, the maximum amounts of specific labelling are in the same ratio in the membranes as for the cells: 0, 1 and 4 for pJDB207, FL442-4B and pAB respectively. As previously observed for labelled cells (Fig. 6A), a polypeptide of about 45 kDa was observed on SDSjPAGE for pAB strain when labelling was performed directly on membrane preparations whatever the N3[3H]Ace concentration used (0.5 - 50 pM). In the case of the membranes isolated from the wild-type strain, a faint but significant band could be detected only when labelling was done with azidoadenine of the highest specific radioactivity available. The molecular mass of this

60

80

(PW

Fig. 7. Specific labelling of plasma membrane preparations N3[3H]Ade. Plasma-membrane-enriched fractions from pAB (m), FL442-4B ( 0 ) and pJDB207 (A) strains were resuspended to 1 mg/ml in 50 mM sodium citrate pH 5.5 containing 2% glucose, 100 mM NH4CI and different concentrations of N3[3H]Ade(0 - 100 pM). After irradiation at 4”C, samples were washed and assayed for protein and radioactivity. Specific labelling was determined as the difference between total labelling (measured in the absence of cytosine) and non-specific labelling (measured in the presence of 2 mM cytosine). As already observed for the cells, non-specific labelling increased linearly with N3[3H]Ade concentration with a slope of 70- 100 pmol (mg protein)-’ (pM N3[3H]Ade)-’ and was quite similar for membranes prepared from the three strains

polypeptide was also in the 45 - 50-kDa region. To visualize this labelling better, SDSjPAGE slab gels at different N3[3H]Ade concentrations were analysed by fluorography at different exposure lengths (Fig. 8). After a 10-day exposure, one spot corresponding to a polypeptide of 40-45 kDa was detected only for the strain with amplified FCY2 gene (Fig. 8A, lane 3). If the exposure exceeded three weeks, an additional less distinct band was observed in the 50-kDa region (Fig. 8B, lane 2’). Under these conditions, samples from the wild-type strain displayed two distinct bands of 45 kDa and 50 kDa (Fig. 8B, lane 5’). Again, nothing was detected for pJDB207 membranes (Fig. 8 A, lane 9). These results show that the plasmid expression in pAB cells leads to an overproduction of a 45-kDa polypeptide, and that both the 45-kDa and 50-kDa polypeptides are probably related to the permease system, since they have not been detected for pJDB207. What is not clear from these results is the relationship between these two peptides. At this stage of our studies, no clear explanation can be offered. Of course, it cannot be totally excluded that proteolysis of the 50-kDa protein might occur during isolation of the membranes, despite the presence of protease inhibitors in the buffers. DISCUSSION 8-Azidoadenine is an active-site-directed photoaffinity label of purine -cytosine permease for the following reasons : (a) it is a potent competitive inhibitor of cytosine uptake (Ki = 3 pM); (b) the covalent binding of the nitrene radical (created upon irradiation) onto the cells induces inactivation of cytosine uptake; (c) specific labelling is prevented by adding an excess of substrate (cytosine, adenine) and (d) it labels only the outer face of the plasma membrane, since it is not accumulated at all by the cells. A 3 - 4-fold increase in the amount of covalent specific labelling was obtained for pAB as compared to FL442-4B.

298 5 6 -7 -8 -9 1-0 -1 2_3 -4 - --

A

k Da

- 180 - 116 -- 84 58 48

I

I

-I-

M

M U

FL442 pAB

FL442

pA0 pJDB

B

36.5 26.6

1 ’ 2’3’ 4’ 5’6’ -----

kDa

- 180 - 116 - 84 -- 48 58

--

36.5 26.6

pA0 FL442 FL442 Fig. 8. Fluorograpphy of SDSjPAGE of membrane preparations labelled by N,L3H]Ade. Membrane proteins were labelled as described with N3[31i]Ade (0.2-0.5 pM) of the highest specific radioactivity (850 GBq/mmol). (A) Labelled proteins (40 pg) were subjected to SDSiPAGE. Gels were soaked for 30 min in Amplifier and exposed for fluorography for 10 days (lanes 1-4) or 21 days (lanes 5-10). Lanes 1. 2, 5. 6, samples from FL442-4B; lanes 3, 4, 7, 8, samples from pAB; lanes 9,10, samples from pJDB207,Odd-numbered lanes, samples irradiated in the absence of cytosine (total labelling); evennumbered lanes, samples irradiated in presence of 2 mM cytosine (non-specific labelling). Specific labelling was 13 pmol/mg for pAB, 4 pmol/mg for FL442-4B and 0 for pJDB207. (B) The samples were treated as in A with exposure time of 20 days. Lanes l’, 2’, 40 pg membrane protein of pAB with 5 pmol/mg specific labelling; lanes 3’, 4 , 40 pg membrane protein of FL442-4B with 1 pmol/mg specific labelling; lanes 5’. 6’, 70 pg membrane protein of FL442-4B with 5 pmol/mg specific labelling. Lanes 2’, 3’, 5’, total labelling; lanes l’, 4’, 6’, non-specific labelling

This is consistant with the 5-fold increase of the V , of uptake, due to the amplification of the FCY2 gene in pAB cells. The specific labelling was related to the inactivation of the permease; thus, it may be assumed that the reaction occurred on the transporter molecule itself. Assuming that one molecule of azido derivative is bound to one molecule of 45-kDa polypeptide, it may be estimated that there are about 6 x 10‘ molecules of this peptide/pAB cell.

Two polypeptides with apparent molecular mass on SDS/ PAGE of 45 kDa and 50 kDa were found in both pAB and FL442-4B strains. These two polypeptides were clearly related to the permease since they were not detected in the permeasenull pJDB207 strain, and since they were not labelled in the presence of an excess of permease ligand. The amplification of the FCY2 gene induced a dramatic increase in the amount of the 45-kDa polypeptide which could be the permease itself. Moreover in pAB strain, it could represent about 4% of the total proteins in membrane preparations. The identity of the 50-kDa polypeptide remains to be established. Purification after solubilisation of these polypeptides labelled with N3[3H]Ade, and utilization of antibodies raised against synthetic peptides characteristic of parts of the permease as deduced from the nucleotide sequence of the gene, should allow better characterization of these entities. These two approaches are currently underway in the laboratory. It has to be pointed out that under our experimental conditions, no specific labelling appears to be related to a 110kDa polypeptide for pAB cells. This result does not agree with that obtained by Schmidt et al. [8]. From labelling experiments performed with the same photolabelling agent, these authors have reported that the permease was a glycoprotein of 110 kDa. This discrepancy is most likely explained by differences in the labelling and SDSjPAGE conditions. In particular, careful consideration must be given to the possibility of an apparent ‘specific reaction’ occurring in the absence of NH4C1 in the irradiation buffer, but partially protected by an excess of cytosine or molecules bearing a primary amino group (Chirio, M. C. and Bri.thes, D., unpublished results). In this paper, we present the reproducible specific labelling of a polypeptide of the S. cerevisiae plasma membrane involved in cytosine transport. Since the specific incorporation of the photolabelling agent on a 45-kDa peptide is correlated with inactivation of the cytosine uptake, we postulate that the 45-kDa polypeptide is the purine -cytosine permease. The difference observed between the apparent molecular mass, obtained by SDS/PAGE (45 kDa), and the molecular mass predicted from the sequence of the DNA segment encoding for the FCY2 protein (58 kDa) [7] might reflect post-traductional maturation of the gene product. However, as previously mentioned, proteolysis during membrane preparation cannot be fully excluded and the influence of SDS on molecular mass determination of integral membrane proteins should be considered. We gratefully acknowledge Dr M.R. Chevallier (Institut de Biologic Molkculaire et Cellulaire, Strasbourg) for the gift of mutant strains, advice and helpful discussions. We also acknowledge Prof. 0. Viratelle for constructive review of the manuscript, C. Sarger for skilful technical assistance, M. L. Grellety and R. Cooke for improving the language. This work was supported by the Centre National de la Recherche Scientifique (LP 8231), by the Universiti?de Bordeaux 2 and by a grant from the Conseil GPnCral d’Aquitaine.

REFERENCES 1. Jund, R., Weber, E. & Che vdkr, M. R . (1988) Eur. J . Biochem. 171,417-424. 2. Polack, A. M. & Grenson, M. (1973) Eur. 1.Biochem. 32, 276282. 3. Reichert, U. & Winter, M. (1974) Biochirn. Biophys. Acta 356, 108- 116. 4. Chevallier, M. R., Jund, R. & Lacroute, F. (1975) J . Bacteriol. 12,629 - 641. 5. Reichert, U., Schmidt, R. & Forkt, M. (1975) FEBS Lett. 52, 100- 102.

299 6. Reichert, U. & Forst, M. (1977) FEBS Lett. 83, 325-328. 7. Weber, E., Rodriguez, C., Chevallier, M. R. & Jund, R. (1990) Mol. Microbiol. 4, 585 - 596. 8. Schmidt, R., Manolson, M. F. & Chevallier, M. R. (1984) Proc. Natl Acad. Sci. USA 81, 6276 - 6280. 9. Schmidt, R., Manolson, M. F., Angelides, K. J. & Poole, R. J. (1981) FEBS Lett. 129, 305-308. 10. Maliarik, M. J. & Goldstein, I. J. (1988) J. Biol. Chem. 263, 11274- 11 279. 11. Jund, R. & Lacroute, F. (1970) J . Bacteriol. 102, 605-607. 12. Beggs, J. (1978) Nature 275, 104-109. 13. HaIey, I). E. (1977) Methods Enzymol. 46, 339-346. 14. Valeins, H., Volker, T., Viratelle, 0. & Labouesse, J. (1988) FEBS Lett. 226, 331 -336.

15. Goffeau, A. & Dufour, J. P. (1988) Methods Enzymol. 157,528533. 16. Lang, B., Burger, G., Doxiadis, I., Thomas, Y., Bandlow, W. & Kaudewitz, T. (1977) Anal. Biochem. 77, 110- 121. 17. Franzusoff, A. J. & Cirillo, V. P. (1983) J. Biol. Chem. 258,3608 3614. 18. Ahlers, J. (1984) Can. J. Biochem. Cell. Biol. 62, 998-1005. 19. Laemmli, U. K. (1970) Nature 227, 680-685. 20. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J. (1951) J . Biol. Chem. 193, 265-215. 21. Ottolenghi, P. (1975) Biochem. J. 151,61-66.