Current Genetics

Curr Genet (1990)18:155-160

9 Springer--Verlag 1990

Translational accuracy and sexual differentiation

in Chlamydomonas reinhardtii L. Bult~ and P. Bennoun Service de Photosynth6se, Institut de Biologie Physico-Chimique, 13 rue Pierre et Marie Curie, F-75005 Paris, France Received September 21, 1989/January 22, 1990

Introduction

Summary. A renewal of ribosomes has been previously reported to occur during gametogenesis in C. reinhardtii. In order to further characterize these new ribosomes, we performed pulse-labelling experiments on whole cells of C. reinhardtii, during gametogenesis and in the presence of various aminoglycosides known to alter translational accuracy: Hygromycin and Paromomycin are assumed to increase the rate of translational errors at the level of 80S and 70S ribosomes whereas Kasugamycin is assumed to induce the opposite effect. Three lines of evidence support an increased inaccuracy in protein translation during gametogenesis: (1) gamete cells displayed a higher sensitivity than vegetative cells to Hygromycin and Paromomycin; 4 gg/ml Hygromycin cancelled cytoplasmic protein synthesis in gametes but not in vegetative cells; Paromomycin induced the synthesis of new polypeptides of high molecular weight and of nuclear origin in gametes but not in vegetative cells. In addition, chloroplast protein synthesis was more sensitive to Hygromycin and Paromomycin in gametes than in vegetative cells. (2) Kasugamycin-sensitive alterations of thylakoid membranes were detected during gametogenesis. (3) 35S-misincorporation in the OEE3 polypeptide, of nuclear origin and normally devoid of sulphur containing amino acids, was more than three times higher in gametes than in vegetative cells. This increase was prevented by Kasugamycin, suggesting that 80S translation in gametes was more inaccurate than in vegetative cells. The possible significance of these changes occurring during gametogenic differentiation is discussed in light of the importance of a modulation of translational accuracy at particular stages of the life cycle in other lower eukaryotes.

The unicellular alga C. reinhardtii displays a sexual life cycle which is easy to control under laboratory conditions. Studies on gametogenesis are facilitated because this process is induced synchronously by nitrogen starvation. Although morphological changes are well documented (Martin and Goodenough 1975), little is known about metabolic events occurring during gametogenic differentiation. Studies by Chiang and colleagues have demonstrated the occurrence of a renewal of 80S (cytoplasmic) and 70S (chloroplast) ribosomes during gametogenesis (Siersma and Chiang 1971; Martin et al. 1976). The new 80S ribosomes display specific structural modifications and differ, in vitro, in their susceptibilty towards Hygromycin B (Picard-Bennoun and Bennoun 1985), an aminoglycoside known to increase translational errors. It was previously reported that mutations affecting the sensitivity towards aminoglycosides most often altered the error rate of translation (for reviews see Tuite 1987; Kieu-Ngoc and Coppin-Raynal 1988). In the present paper, we have investigated the aminoglycoside [Paromomycin (Pro), Hygromycin B (Hm) and Kasugamycin (Ks)]-sensitivity in vivo of gamete ribosomes compared to those of vegetative cells: Hm and Pm are usually assumed to decrease the accuracy of translation (for a review see Tuite 1987) whereas Ks would induce the opposite effect (van Buul et al. 1984). We attempted to evaluate whether a modulation of the rate of translational errors occurred in the course of the life cycle of C.

Key words: Chlamydomonas reinhardtii - Gametogenesis - Accuracy of translation Aminoglycosides

Material and methods

Offprint requests to." L. Bult6 Abbreviations: Hm: Hygromycin B; Pro: Paromomycin; Ks: Kasugamycin; CAP: Chloramphenicol; OEE: Oxygen evolving enhancer

reinhardtii.

Strains and culture conditions. Vegetative cells of the wild type of C. reinhardtii (strain 137c) were grown at 27~ in tris-acetate-phosphate medium (Gorman and Levine 1965) under an illumination of 500 lux. The C4 mutant, isolated by M. Delosme and P. Bennoun, is unable to produce sexually competent cells when deprived of nitrogen sources. The FUD50 mutant lacks the chloroplast ATP synthase (Woessner et al. 1984). To induce gametogenesis, one vol-

t56 ume of vegetative celIs (8.106 cells/ml) was diluted with four volumes of minimal medium supplemented with acetate in which nitrogen was omitted, and stirred under 450 lux at 27~ Under these conditions, vegetative cells underwent a synchronous division into four cells, starting about 15 h after nitrogen starvation: four gametic cell-figures appeared that were dissociated 5 h later; at about 28 h, these cells acquired a sexual competence which was maintained for several days. Chemicals. Pm was a gift of Substantia Laboratoires (Courbevoie, France). Hm, Ks, CAP and cycloheximide were purchased from Sigma Inc (St Louis, USA). The effect of HM, Pm, Ks and cycloheximide on vegetative growth was evaluated by the increased population of cells grown under continuous 500 lux light in TAP medium. The effect of Hm, Pm and Ks on sexual differentiation was evaluated by the ability of gamete cells of mating type + and , differentiated in the presence of various concentrations of these drugs, to fuse within 2 h at 1000 lux. The percentage of cells unable to mate was estimated through the number of swimming cells. Assays of in vivo protein synthesis in vegetative and gamete cells. 14C-acetate (2 gCi/ml) - and 35SO4 (5 laCi/ml) - pulse-labelling of whole cells were performed under an illumination of 500 lux for 45 rain in acetate-free medium and 120 rain in SO~-free medium, respectively (Delepelaire 1983). In the latter case, gamete cells, like vegetative cells, were washed and resuspended, prior to pulselabelling, in a medium supplemented with nitrogen, in order to eliminate a possible effect induced by amino-acid starvation. Chloroplast protein synthesis is defined by the remaining protein synthesis in the presence of 8 p.g/ml cycloheximide, an inhibitor of 80S ribosomes. Cytoplasmic protein synthesis is the protein synthesis detected in the presence of 100 gg/ml CAP, an inhibitor of 70S ribosomes (Delepelaire 1984). Biochemical analysis of thylakoid membranes. Purified thylakoid membranes were isolated according to Chua and Bennoun (1975). The extrinsic OEE3 subunit was extracted by alkaline treatment, which releases extrinsic polypeptides (Rousselet and Wollman 1986), using FUD50 thylakoid membranes (1 mg chlorophyll/ml) in order to avoid release of the ~ subunit of CF1 whose electrophoretic mobility is similar to that of the OEE3 subunit (Lemaire and Wollman 1989). Polypeptides released in the supernatant (300000 g for 15 min) were solubilized in the presence of 1% SDS. SDS/urea electrophoresis was run as described by Delepelaire (1983). Polypeptides were either stained by Coomassie blue or silver stained. Autoradiography of the dried gels was performed using Agfa-Gevaert industrial P films. Densitometric scanning of the autoradiograms involved the use of a LKB Ultroscan densitometer associated with LKB 2400 Gel Scan XL software. For estimation of cytoplasmic protein synthesis, a significant number of nuclear polypeptides were chosen by densitometric scanning of the autoradiogram produced by the electrophoresis of the total polypeptide content of pulse-labelled cells.

Results F o r convenience, we defined four cytological stages characteristic o f gametogenesis: (1) t0, the initial step, when cells were transferred into nitrogen-free m e d i u m (2) tl ( a b o u t 15 h after tO): the beginning o f the f o r m a t i o n o f four cell-figures (3) t2 (about 20 h after tO) when these cells were free (4) t3 (about 28 h after tO) when the free cells became c o m p e t e n t gametes (able to mate). The analysis o f thylakoid m e m b r a n e s o f t3 m a t u r e gametes after S D S - P A G E revealed a decreased content in

Fig. 1. a Enlargement of the cyt.f region (arrow) after urea SDS-gel electrophoresis of thylakoid membranes of vegetative cells (tO) and gamete cells at t2 and t3 differentiated in the absence (t3) or presence (t3 +Ks) or 1 mg/ml Ks. b Enlargement of the 23-30 K region of the autoradiogram corresponding to a 7.5-15% polyacrylamide SDS gel loaded with thylakoid membranes from 14C-acetate pulselabelled WT vegetive cells at tO and gamete cells at t3 differentiated in the absence (t3) or presence (t3 + Ks) of 1 mg/ml Ks; similar experiments with the C4 mutant unable to produce sexually compentent cells: cells were at tO and after 40 h of nitrogen-starvation ("t3")

several complexes o f the p h o t o s y n t h e t i c chain. One early difference consisted in the gradual disappearance o f the t r a n s m e m b r a n e cyt.b6/f complex. This decay became detectable at t2 (exemplified by the disappearance o f cyt.f in Fig. 1 a, f r o m tO to t2 and t3). The analysis o f radiolabelled 14C-incorporation in thylakoid m e m b r a n e polypeptides during 1 h, indicated a deficiency in the synthesis o f two polypeptides o f the light harvesting complex (numbers 15 and 17) at t3 (Fig. 1 b, lane t3). The C4 m u t a n t , which is unable to mate, did n o t display a n y o f the a b o v e described alterations when starved o f nitrogen sources. This is depicted in Fig. 1 b, lanes tO and "t3", by the absence o f L H C deficiencies in this mutant. Interestingly, Ks, assumed to increase translational accuracy, almost completely prevented the disappearance o f cyt.b6/f in the thylakoid m e m b r a n e s o f gametes (Fig. 1 a, lane t3 + Ks and see Bult~ and W o l l m a n 1990). Nevertheless, Ks failed to restore the synthesis o f the two polypeptides (15 and 17) o f the light harvesting complex which are deficient in gametes (Fig. I b, lane t3 + Ks). These observations suggest that a m o d u l a t i o n o f translational accuracy during gametogenesis m a y be responsible for some o f the polypeptide changes which occur in thylakoid membranes. We therefore attempted to investigate whether there was an increased sensitivity, in vivo, towards H m , P m and Ks, which display opposite effects on the accuracy o f translation at different stages o f the life cycle o f C. reinhardtii. The effect o f these aminoglycosides on the accuracy o f translation was c o m p a r e d with the effect o f cycloheximide which does n o t m o d i f y the accuracy o f translation (Tuite 1987).

157 Table 1. Effect of rim, Pm and Ks on vegetative growth and mating

Vegetative growth and gametic differentiation Table 1 shows the minimal concentration of Pm and H m able to cancel vegetative growth and mating in our culture conditions (see Material and methods). Cycloheximide, employed at 0.125 gg/ml, a concentration that produced the same rate of inhibition of protein synthesis as 4 gg/ml H m (Fig. 2, left part, tO), slowed down growth by only 50%, i.e., to the same extent as it reduced 80S protein synthesis. These observations suggested a particular effect of H m and Pm which would not be linked to an inhibition of elongation, but, rather, to their possible effect on the accuracy of translation. Similarly, I mg/ml Ks had a 50% inhibition effect on vegetative growth, a concentration where no inhibition of 80S protein synthesis was detected in pulse-labelling experiments (data not shown). In addition, the minimal concentrations of H m and Pm inhibiting vegetative growth did not prevent gametogenesis. Ks up to 1 mg/ml did not prevent gametogenesis, although it modified some of the phenotypical features displayed by mature gametes (see Fig. 1 a, t3 + Ks).

LCI a of growth LCI of mating b

Hm

Pm

Ks

2 4

1 2

> 1000 > 1000

LCI: Lower concentration for total inhibition (gg/ml) b Hm, Pm and Ks were added to the differentiation medium at tO and their effects were evaluated through the ability of cells to mate with gametes of the opposite mating type whether or not treated with the same antibiotic

Protein synthesis." effect of i l m and Pm on cytoplasmic protein synthesis H m (4 gg/ml) inhibited cytoplasmic protein synthesis by only 50% at tO whereas it totally blocked cytoplasmic protein synthesis at t3 (Fig. 2 left, lanes 2), as did saturating concentrations (8 gg/ml) of cycloheximide. This sensitivity towards H m altered concurrently with the main stages ofgametogenesis (Fig. 3): similar to that of vegetative cells during the first steps of differentiation (tO and tl), the hypersensitivity was almost completely acquired at t2. This hypersensitivity could have several origins: (1) an increase in the antibiotic/ribosome ratio, resulting from a reduction in the number of ribosomes per cell during the course of gametogenesis (Martin et al. 1976). (2) an increase in the sensitivity of ribosomes to inhibitors of translation. (3) an effect of nitrogen starvation. Hypothesis (1) can be excluded since the level of inhibition by H m remained constant when varying the concentration of vegetative cells (2.106 c/ml to 107 c/ml) or gamete cells (4.106 to 4.10 v c/ml) during the pulse-labelling period (data not shown). Thus variations in ribosome concentration had no significant effect on the inhibition by H m (4 gg/ml). We determined the concentration of cycloheximide producing the same inhibition of 80S protein synthesis as H m (4 gg/ml) does at tO and observed that this rate of inhibition did not change during gametogenesis (Fig. 2 left, lane 2 at tO and t3). This rules out hypothesis (2). The C4 mutant did not gain any hypersensitivity towards Hygromycin after nitrogen deprivation for 40 h; the sensitivity of t3 gametes towards H m (4 gg/ml) was the same whether or not they were pulse-labelled in the presence of nitrogen sources (data not shown). Therefore, an effect of nitrogen starvation is unlikely. We thus

Fig. 2. Similar experiment as in Fig. I b showing the total protein content in whole cells at tO and t3. Left part: effect of Hm: in the presence of 100 gg/ml chloramphenicol without (lane 1) or with (lane 2) 4 gg/ml Hm, or with 0.125 gg/ml cyctoheximide (lane 3). Right part: effect of Pro: without protein synthesis inhibitors (lane 1) and with 4 gg/ml Pm (lane 2). R and R' indicate polypeptides specifically synthesized in the presence of Pm

z 5O

A

o9 I.IJ NZ

25 Z~

Z

L/A t-.-

or r

to

J

tl

t2

t3

Fig. 3. Relative amount of cytoplasmic protein synthesis remaining in the presence of Hm during gametogenesis. Cells were pulse-labelled during 45 rain at different times, ti, of differentiation, in the presence of CAP 100 gg/ml and in the presence or the absence of Hm 4 gg/ml. 14C incorporation in nuclear polypeptides was calculated from the ratio of labelling with/without Hm. For details see Material and methods

158 conclude that the hypersensitivity of cytoplasmic protein synthesis towards Hm in vivo probably originates from a specific property of the newly synthesized ribosomes which would appear, during gametogenesis, between tl and t2. The analysis of the action of Pm gave qualitatively different results. At t2, as well as at t3, the synthesis of most cytoplasmically-translated polypeptides decreased significantly whereas new polypeptides appeared, mainly in the high molecular weight region (R and R' in Fig. 2 right and Fig. 4, t0 and t2, lanes 4). These were not observed with Hm (data not shown). Their nuclear origin is likely since they did not appear when cells were pulse-labelled in the presence of cycloheximide (8 pg/ml; Fig. 4, t2, lanes 5 and 6). We then compared the increasing inhibition of cytoplasmic protein synthesis at tO and t2 in parallel with the appearance of R and as a function of increasing concentrations of Pm (Fig. 5). Cytoplasmic protein synthesis remained slightly more sensitive to Pm at t2 (50% inhibition: 2 gg/ml) than at tO (50% inhibition: 4 ~tg/ml) and the R polypeptide appeared at lower Pm concentrations, and in greater amounts, at t2 (with a maximum at 2 ~tg/ml) than at tO (with a maximum at 3 gg/ml). This could be explained by the ability of Pm to increase translational errors and to induce the synthesis of readthrough products at the expense of conventional protein synthesis (Zierhut et al. 1979; Tuite and McLaughlin 1984).

Protein synthesis." effect o f rim and P m on chloroplast protein synthesis

A typical pattern of chloroplast protein synthesis is shown in the first lane of Fig. 6 obtained by pulse-labelling cells in the presence of saturating concentrations of cycloheximide, an inhibitor of cytoplasmic protein synthesis. Pulse labelling at tO and t2 revealed a great resistance of chloroplast protein synthesis towards high concentrations of Hm and Pm. Additionally, Hm (10 gg/ml) produced the same effect as cyclo (8 ~tg/ml): there was no detectable inhibition of chloroplast protein synthesis whereas cytoplasmic protein synthesis was cancelled at this concentration (Fig. 6, lanes 2). These experiments do not allow a discrimination between a resistance of 70S ribosomes and a lack of permeation of the chloroplast envelope to Hm or Pm; nevertheless, concentrations higher than 10 gg/ml Hm (lanes 3) inhibited chloroplast protein synthesis at t2 to a greater extent than at tO. In contrast, many polypeptides of cytoplasmic origin (as defined by their sensitivity towards cycloheximide), were still synthesized in cells pulse-labelled in the presence of Pm (Fig. 6, lanes 4, 5, 6) whereas the synthesis of chloroplast polypeptides was partially inhibited. Nevertheless, chloroplast protein synthesis, exemplified by three easily detectable polypeptides (and designed with a star in Fig. 6), was much more sensitive to Pm at t2 than at tO. Thus, as with the 80S cytoplasmic ribosomes, the newly synthesized 70S chloroplast ribosomes displayed an increased sensitivity towards Hm and Pm as compared to their vegetative counterparts.

Protein synthesis: effect o f Ks

A concentration of 1 mg/ml Ks produced no detectable inhibition of cytoplasmic or chloroplast protein synthesis (data not shown). However, since Ks prevented some of the modifications in gamete phenotype (see first section), we wondered whether gametes differentiated in the presence of Ks displayed the same hypersensitivity towards Hm as untreated gametes. We observed the same hypersensitivity towards Hm under these conditions (data not shown). Thus we could rule out an action of Ks during the assembly of the new ribosomes.

Fig. 4. Effectof increasing concentrations of Pm on ~4C incorporation in whole cells polypeptides at tO and t2 (autoradiogram of an SDS-urea gel). Lanes O, 1, 2, 3, 4.' 0, 1, 2, 3, 4 gg/ml Pm. R and R' correspond to polypeptides specificallysynthesized in the presence of Pm. Lanes 5 and 6: cells, pulse-labelled at t2, in the presence of cycloheximide 8 gg/ml and in the absence (lane 5) or presence (lane 6) of Pm 4 gg/ml

v ff) o3

I ~Z

5O -%

Z nl

o~r IX.

1

2

3

4

1~O

")) ........... 4 0

Fig. 5. Effect of increasing concentrations of Pm in cells at tO (o) and t2 (A) on cytoplasmic protein synthesis (dashed lines) and on the synthesis of R polypeptide (solid lines) evaluated through the scanning of autoradiograms of Fig. 5. 14C incorporation in nuclear polypeptides in the presence of Pm is normalised to 1~C incorporation without Pro; the rate of R synthesis is estimated by the amount of 14C incorporation in R over the overall incorporation in whole cells without Pm: the maximal rate of R synthesis is 5.5 times higher at t2 than at tO

159

Fig. 6. Effect of Hm and Pm on chloroplast protein synthesisin whole cells at tO and t2: (lane 1) no inhibitors added. (Lanes 2 and 3) 10 and 100 gg/mlHm. (Lanes 4, 5, 6) 4, 10, 40 lag/mlPm. The first lane displaysa conventionalpattern of chloroplastprotein synthesis obtained by pulse-labellingthe cellsin the presenceof 8 gg/mlcycloheximide(eyclo); the star points out three easilydetectablepolypeptides, of chloroplast origin; Hm 10 Ixg/mlproduces an effectsimilar to cycloheximide

corporation and misincorporation. The autoradiogram (lane 2) reveals a very weak labelling of the OEE3 polypeptide, corresponding to a misincorporation. This misincorporation should be related to the activity of ribosomes during the pulse-labelling period that might be different in vegetative cells and gamete cells, whether or not differentiated, in the presence of Ks. We therefore chose, among other alkaline-sensitive polypeptides, three which showed high labelling, although they were present in much lower amounts when viewed after Coomassie blue staining (see polypeptides al, a2, a3); consequently, 3ss incorporation in these polypeptides is very weakly sensitive to slight variations due to a modulation of the misincorporation level. The incorporation of asS in OEE3 polypeptide in each sample could thus be normalized to the incorporation of 35S in al, a2 and a3. The means of the resulting values (2; OEE3/ai) were normalised for each sample to the same mean obtained in vegetative cells. 35S incorporation in OEE3 was 3.5 times higher in gamete cells than in vegetative cells. This observation is consistent with an increased inaccuracy of the 80S gamete ribosomes in reading the OEE3 messenger. On the other hand, we observed that Ks decreased this misincorporation in gametes by two-fold, which suggests that this antibiotic decreased the error level rate of 80S translation in C. reinhardtii.

Discussion

Fig. 7. See text

Sulphur amino acid misincorporation in the thylakoid polypeptide O EE3

The OEE3 polypeptide sequence contains no sulphur amino acids (Mayfield et al. 1989); thus, the analysis of the misincorporation of sulphur-containing amino acids into this polypeptide could provide a more direct estimation of the error level of translation. To this end, vegetative cells and gamete cells (t3) whether differentiated with Ks or not, were pulse-labelled with 35S sulphate. An alkaline wash of labelled thylakoid membranes released the OEE3 subunit among other extrinsic polypeptides (Fig. 7, lane 1). Total incorporation in a given polypeptide includes normal sulphur-containing amino acid in-

In this report, we have demonstrated in vivo that both 80S and 70S translation acquire an altered sensitivity towards Hm and Pm during gametogenesis of C. reinhardtii. These two aminoglycosides display a dual effect on protein synthesis when bound to ribosomes: they have also been demonstrated to be strong promoters of misreading in prokaryotic as well as in eukaryotic systems (for a review see Tuite 1987) in addition to having a strong inhibitory effect on the elongation stage of translation (Cabanas et al. 1978). The result of this dual action will depend on whether or not a significant increase in the level of misreading occurs before the inhibition of protein synthesis, In most cases, mutants characterized by an altered sensitivity towards Hm or Pm were shown to display an altered accuracy of translation (Tuite 1987). In C. reinhardtii, previous observations indicated the occurrence during gametogenesis of a renewal of both cytoplasmic and chloroplast ribosomes (Siersrna and Chiang 1971) and of structural modifications in the 80S ribosomes (Picard-Bennoun and Bennoun 1985). We observed that 70S and 80S translation in gametes displays a hypersensitivity, in vivo, towards Hm and Pro. Interestingly, Pm, but not Hm, induced the synthesis of new cytoplasmic polypeptides of high apparent molecular weight; these new polypeptides occured m gamete cells to a much greater extent, and at lower Pm concentrations, than in vegetative cells. They may represent translational misproducts since in vitro studies on 70S ribosomes (Zierhut et al. 1979) showed that Pm induces detectable levels of readthrough products whereas Hm does not.

160 These results were obtained using an experimental procedure in which cells were pulse-labelled in minimal medium (devoid of acetate) using 14C acetate as a marker of protein synthesis; interestingly, in the presence of acetate (using 35S as a marker), we observed both Pm- and Hm-induced cytoplasmic "translational misproducts" in gamete cells but not in vegetative cells. Under these conditions, and in contrast with experiments in the presence of 14C acetate, cytoplasmic protein synthesis is more sensitive to these aminoglycosides in vegetative than in gamete cells (unpublished observations). As pointed out previously (Laughrea et al. 1984) the action of drugs in inducing translational errors is modulated by cofactors such as Mg + +. Therefore, a modification of the medium composition could modulate the action of Pro and Hm; in any case, gamete and vegetative 80S translation display distinctive responses, in vivo, towards these two aminoglycosides, which favours the hypothesis that the structural changes observed in the 80S gamete ribosomes affect translational accuracy. More direct evidence of an increase in the rate of translational errors was provided by the analysis of misincorporation of sulphur amino acids in the thylakoid polypeptide OEE3 of nuclear origin: this misincorporation is increased more than three times in gametes as compared to vegetative cells; in addition it is reduced two fold by Ks, a ribosome-directed aminoglycoside which, in contrast with others, increases the accuracy of translation in bacterial ribosomes (Van Buul et al. 1984). Indirect evidence suggests that it produces the same effects on eukaryotic ribosomes (Kieu-Ngoc and Coppin-Raynal 1988). Thus, our data favour the conclusion that cytoplasmic protein synthesis is less accurate in gametes than in vegetative cells; moreover, they indicate that Ks increases translational accuracy in C. reinhardtii, as it does in other organisms. We also observed specific changes in the thylakoid transmembrane complexes of the photosynthetic apparatus in gamete cells. Indeed, equivalent biochemical alterations also occur in the course of gametogenesis, when new ribosomes appear, as detected by their altered sensitivity in vivo towards H m and Pm. We showed that Ks prevents some of the alterations in the thylakoid membranes, which is consistent with its ability to restore higher translational accuracy. These alterations would thus originate from an increased ambiguity of cytoplasmic translation in gamete cells. The exact function of the synthesis of new ribosomes with altered translational accuracy during the life cycle of C. reinhardtii remains unclear. We noticed that in the C4 mutant, which has impaired gametogenesis, nitrogen starvation neither induces alteration of translational accuracy, as detected by ribosome sensitivity towards H m and Pro, nor thylakoid membrane modifications. It is thus possible that the aquisition of sexual competence requires an increase in translational errors, at particular stages. It may also be that gamete ribosomes play an important role in subsequent phases of the sexual cycle.

It has been proposed that the rate of translational errors is modulated in other lower eukaryotes: in Podospora anserina, readthrough or frameshift errors may allow the synthesis of small amounts of regulatory proteins, essential for sporulation to proceed (Coppin-Raynal et al. 1988, Dequard-Chablat and Coppin-Raynal 1984; PicardBennoun et al. 1983). Sporulation in yeast may also depend upon a particular change in translational fidelity (Rothstein et al. 1977).

Acknowledgements. We thank Y. Pierre and M. Delosme for excellent technical assistance, P. Delepelaire for fruitful discussions, E-A. Wollman whose advice and critical reading of the manuscript were essential in this work, and Ms Couratier for the micrographs. This work was supported by the C.N.R.S. (U.A.1187). L. Bult6 is a recipient of a B.D.I.C.N.R.S. fellowship.

References

BuR+ L, Wollman F-A (1990) Proceedings of the VIIP h international congress of photosynthesis. Kluwer, Dordrecht Boston London, pp 13-15 Cabanas MJ, Vasquez D, Modolell J (1978) Eur J Biochem 87: 21' 27 Chua NH, Bennoun P (1975) Proc Natl Acad Sci USA 72: 21752179 Coppin-Raynal E, Dequard-Chablat M, Picard M (1988) In: Tuite MF, Picard M, Bolotin-Fukuhara M (eds) Genetics of translation. Springer, Berlin Heidelberg, pp 431-442 Delepelaire P (1983) Photobiochem Photobiophys 6:279-291 Delepelaire P (1984) EMBO J 3:701-706 Dequard-Chablat M, Coppin-Raynal E (1984) Mol Gen Genet 195:294 299 Gorman DS, Levine RP (1965) Proc Natl Acad Sci USA 54: 16651669 Kieu-Ngoc A, Coppin-Raynal E (1988) Genet Res 88:179-184 Laughrea M, Latulippe J, Filion AM (1984) Biochem 23:753 758 Lemaire C, Wollman F-A (1989) J Biol Chem 264:10228-10234 Mayfield SP, Schirmer-Rahire M, Frank G, Zuber H, Rochaix JD (1989) Plant Mol Biol 12:683-693 Picard-Bennoun M, Bennoun P (1985) Curr Genet 9:239-243 Picard-Bennoun M, Coppin-Raynal E, Dequard-Chablat M (1983) In: Abraham et al. (eds) Protein synthesis. Humana Press Inc, pp 221 232 Martin NC, Goodenough VW (1975) J Cell Biol 67:587-605 Martin NC, Chiang KS, Goodenough VW (1976) Dev Biol 51: 190201 Rothstein R J, Esposito RE, Esposito MS (1977) Genetics 85: 35-54 Rousselet A, Wollman F-A (1986) Arch Biochem Biophys 246:321 331 Siersma PW, Chiang KS (1971) J Mol Biol 58:167-185 Tuite MF (1987) In: Rose AH, Harrison JS (eds) The yeasts vol 4. The genetics and Biochemistry of yeast protein synthesis Tuite MF, McLaughlin CS (1984) Biochim Biophys Acta 783: 166170 van Buul CPJJ, Visser W, van Knippenberg PH (1984) FEBS Lett 177:119-123 Woessner JP, Masson A, Harris EH, Bennoun P, Gillham NW, Boynton JE (1984) Plant Mol Biol 3:177-190 Zierhut G, Piepersberg W, B6ck A (1979) Eur J Biochem 98: 577583

Communicated by K. P. Van Winkle-Swift