Eur. J. Biochem. 267, 3226±3234 (2000) q FEBS 2000

Regulation of secretory protein gene expression in Paramecium Role of the cortical exocytotic sites AngeÂlique Galvani and Linda Sperling Centre de GeÂneÂtique MoleÂculaire, CNRS, Gif-sur-Yvette, France

In cells that possess a regulated secretory pathway, exocytosis can lead to transcriptional activation of genes encoding products stored in secretory granules as well as genes required for granule biogenesis. With the objective of understanding this response, we have examined the expression of Paramecium secretory protein genes in different physiological and genetic contexts. The genes belong to the trichocyst matrix protein (TMP) multigene family, encoding polypeptides that form the crystalline matrix of the secretory granules, known as trichocysts. Approximately 1000 trichocysts per cell are docked at pre-formed cortical exocytotic sites. Their rapid and synchronous exocytosis can be triggered by vital secretagogues such as aminoethyldextran without harming the cells. Using this exocytotic trigger, we found that the transcription of TMP genes undergoes rapid, transient and co-ordinate 10-fold activation in response to massive exocytosis, leading to a 2.5-fold increase in the pool of TMP mRNA. Experiments with exocytosis-deficient mutants show that the secretagogue-induced increase in intracellular free calcium implicated in stimulus/secretion coupling is not sufficient to activate TMP gene expression. We present evidence that the state of occupation of the cortical exocytotic sites can affect TMP gene expression and suggest that these sites play a role in gene activation in response to exocytosis. Keywords: exocytosis; Paramecium; secretory mutants; transcription; trichocyst.

Regulated secretion is a function that allows certain eukaryotic cells to store secretory products in specialized organelles and to release them only in response to appropriate extracellular stimulation (reviewed in [1,2]). Data from several systems have shown that recovery from secretory activity can involve the transcriptional activation of genes required for synthesis and maturation of the secretory products as well as their packaging into granules [3±9]. These observations raise the question of whether or not the signals involved in stimulus/secretion coupling are also responsible for the transcriptional activation. Paramecium presents an attractive model for studying this and other aspects of regulated secretion because (a) massive and synchronous degranulation can be obtained without killing the cells, and it is readily evaluated by using simple light microscopy techniques and (b) the secretory function is not vital for growth under laboratory conditions so that the pathway can be dissected genetically (reviewed in [10]). In Paramecium, about 1000 voluminous spindle-shaped secretory vesicles, known as trichocysts, are docked at specific cortical exocytotic sites, as illustrated in Fig. 1 for the wild-type and a mutant in which the trichocysts are unable to dock. Exocytosis in response to extracellular stimulation is extremely rapid: signal reception and transduction, membrane fusion and release of granule contents is complete within 100 ms of triggering [11]. During these events, the exocytotic sites are dynamically remodelled [12±16]. After exocytosis, the trichocyst membrane is internalized [17] and the site returns to its unoccupied configuration and can participate in further rounds Correspondence to L. Sperling, Centre de GeÂneÂtique MoleÂculaire, CNRS, 91198 Gif-sur-Yvette Cedex, France. Fax 1 33 1 69 82 31 50, Tel.: 1 33 1 69 82 32 09, E-mail [email protected] Abbreviations: AED, aminoethyldextran; ICL, infraciliary lattice; TMP, trichocyst matrix protein. (Received 6 March 2000, accepted 28 March 2000)

of secretory activity. Recovery of the cell's exocytotic capacity after complete degranulation, involving the biogenesis of a whole new set of trichocysts, takes up to 9 h [18]. Many genetic loci necessary for trichocyst docking and for the final exocytotic events (signal reception, transduction and membrane fusion) have been identified by isolation and characterization of exocytosis-deficient mutants and four of the genes have been cloned by functional complementation [19] (J. Cohen, personal communication). The gene encoding calmodulin is also required [20]. Availability of this collection of mutants has helped to obtain evidence by biochemical, electrophysiological and calcium imaging techniques that a transient subcortical increase in intracellular free Ca21 is both necessary and sufficient for exocytotic membrane fusion, and involves stimulus-dependent influx of Ca21 from the external medium as well as stimulus-dependent release of Ca21 from internal stores [21±25]. We report here the effects of exocytosis on the expression of genes encoding secretory proteins stored in the trichocysts. These trichocyst matrix protein (TMP) genes belong to a large multigene family whose co-expression generates the heterogeneous mixture of polypeptides that constitute the crystalline trichocyst matrix (see [10] for a review describing the TMP genes and their function in trichocyst biogenesis). We found co-ordinate transcriptional activation of the genes following exocytosis, leading to an increase in the pool of TMP mRNA. We have established that the kinetics of the transcriptional response involves rapid activation and inactivation, using an in vitro nuclear transcription assay. Experiments with secretory mutants show that the exocytosis-associated Ca21 movements are not sufficient for gene activation. However transcription rate and mRNA levels are elevated in mutants with unoccupied trichocyst docking sites, suggesting that signals generated at these cortical exocytotic sites can be transmitted to the nucleus to regulate TMP gene expression.

q FEBS 2000

Expression of Paramecium secretory protein genes (Eur. J. Biochem. 267) 3227

Secretagogue stimulation of mass exocytosis The well characterized vital polycationic secretagogue aminoethyldextran (AED) [28,29] was used to stimulate trichocyst exocytosis. A log-phase culture (< 1000 cells´mL21) was collected by brief centrifugation (500 g), and the pellet was expelled into 100 mL buffer containing AED (kindly provided by J. Cohen) at a concentration of 15 mm in 10 mm Tris/HCl, 1 mm CaCl2, pH 7.0. Wild-type cells, but not the secretory mutants, secreted all of their docked trichocysts, which appeared as clumps of flocculent material in the buffer. Cells were immediately washed three times in fresh medium to remove all traces of secretagogue. The success of the operation was systematically verified using the picric acid test of exocytotic capacity (see below) to show that no secretable trichocysts remained in the cells. Evaluation of exocytotic capacity Fig. 1. Nomarski images of a wild-type and a tam8 mutant cell, showing the localization of the trichocysts (tr). In wild-type cells, the trichocysts are inserted at docking sites located along the ciliary rows, which are only in the plane of focus in this image at the cell periphery (see inset). In tam8 mutant cells the trichocysts cannot attach to the cortex and can be seen, unattached, throughout the cytoplasm. M, Macronucleus; OA, oral apparatus; DV, digestive vacuoles. Wild-type cell anterior is at the top of the page; tam8 mutant cell anterior is at the bottom of the page. The bright spots are inorganic crystals. Bars, 10 mm.

M AT E R I A L S A N D M E T H O D S Cells and culture conditions The wild-type reference strain was Paramecium tetraurelia d4-2, derived from stock 51 [26]. The secretory mutants used in this study, their phenotypes, and the site of action of the mutations (trichocyst, plasma membrane or cytosol) are presented in Table 1. Cells were grown at 27 8C in grass infusion (Wheat Grass Powder, Pines International, Lawrence, KS, USA), bacterized with Klebsiella pneumoniae the day before use, and supplemented with 0.4 mg´mL21 b-sitosterol [27].

The efficiency of mass exocytosis and the time course of subsequent resynthesis were evaluated using picric acid [30], as illustrated in Fig. 2. Cells were transferred to a drop of a saturated solution of picric acid using a micropipette. This fixative induces complete release of exocytosis-competent trichocysts. The discharged material, which remains connected to the cell surface, can easily be visualized under dark-field light microscopy at low magnification. The number of exocytosis competent trichocysts can be either counted (0±50 secreted trichocysts) or estimated from the appearance of the halo, assuming that the dense silky halo observed for wild-type cells represents about 1000 secreted trichocysts. Photomicroscopy Cells treated with picric acid were viewed with a Zeiss axiomat photomicroscope under pseudo-dark field illumination using a 10 objective and photographed with Kodak Tmax 400 film. For Nomarski phase interference contrast microscopy, cells were permeabilized with a digitonin-saturated solution of PHEM buffer (60 mm Pipes, 25 mm Hepes, 10 mm EGTA, 2 mm MgCl2, pH6.9) for 30 min and fixed for 30 min in 1% paraformaldehyde in PHEM buffer. Cells were then mounted in Citifluor (Citifluor Ltd., London, UK) supplemented with 2 mm MgCl2 and 10 mm EGTA and observed with a Leica

Table 1. Secretory mutants. All of the mutants used in this study are exocytosis-deficient. The site of action of the mutations was previously determined by rescue experiments [13,51,54] as being either in the trichocyst compartment, the cytosol or the plasma membrane. Allele names take into account current genetic nomenclature for ciliates [55] and are found with other names in the older literature (nd6-1 is nd6a, nd9-3 is nd9c, tam6-1 is tam6a, tam6-2 is tam6b and nd3-5 is nd3e). Mutant

Phenotype

Site of action

Reference

nd3-5 nd6-1 nd7-1 nd9-3 tam1 tam6-1 tam6-2 tam8 tam10 tam11 tam54

Docked trichocysts Docked trichocysts Docked trichocysts Docked trichocysts Undocked trichocysts Undocked trichocysts Undocked trichocysts Undocked trichocysts Undocked trichocysts Undocked trichocysts Undocked trichocysts

Plasma membrane Plasma membrane Trichocyst Cytosol Trichocyst Plasma membrane Plasma membrane Trichocyst Trichocyst Trichocyst Not determined

[51] [13] [13] [51,52] [30] [51] [51] [53] [51] [51] F. Ruiz, unpublished data

3228 A. Galvani and L. Sperling (Eur. J. Biochem. 267)

Fig. 2. Recovery of wild-type cells after massive exocytosis. After AED treatment, cells were treated with picric acid at different times and photographed under dark-field illumination. After stimulation, cells are almost entirely depleted of secretable trichocysts (AED 5 min). At 20 min, the up to 30 mature trichocysts which were still free in the cytoplasm at the time of exocytosis have docked and can be secreted; the number of secretable trichocysts then increases slowly. Complete recovery requires several hours as reported previously [18]. Bar, 20 mm.

photomicroscope and photographed with Kodak Tmax 400 film. RNA extraction and Northern blots Total RNA was prepared from log-phase cultures according to the method of Chomczynski and Sacchi [31], using the Trizol reagent (Gibco-BRL). The only modification of the standard protocol was that the cells were lysed by vortexing in the presence of glass beads. Total RNA (10 mg per lane) was fractionated on formaldehyde/1.25% agarose gels, and transferred to positively charged nylon membranes (Ambion) by capillarity. Membranes were UV cross-linked and finally washed in 2  NaCl/Cit for 5 min. Filters were prehybridized for 2 h at 60 8C in 6  NaCl/Cit, 2  Denhardt's solution, 0.1% SDS. Hybridizations were carried out at the same temperature during 16±24 h in the same buffer, with 106 c.p.m. 32P-labelled probe per mL hybridization buffer. Membranes were washed for 5 min at room temperature and 30 min at 60 8C, first in 2  NaCl/Cit, 0.1% SDS and then in 0.2  NaCl/Cit, 0.1% SDS. Hybridizations were quantified using a PhosphorImager and imagequant software (Molecular Dynamics). Preparation of radioactive probes for Northern blots The specific probes used in this study correspond to members of multigene families that are organized in subfamilies of very

q FEBS 2000

similar genes (. 85% nucleotide identity) which code for very similar proteins. T1, T2 and T4 are subfamilies of the TMP multigene family [32] of which T1b, T2c and T4a are specific members [33]. Likewise, infraciliary lattice (ICL)1 is a subfamily of Paramecium centrin genes of which ICL1a is a specific member [34]. We consider these probes to be subfamily- rather than gene-specific given the hybridization conditions used here (see [32±34]). The ICL1a centrin gene [34,35] was chosen as a normalization control for the Northern blot experiments as the centrin mRNA is smaller than T1 mRNA allowing simultaneous hybridization of blots with both probes, and given that the amount of centrin mRNA does not appear to vary under the conditions used, and is always proportional to the 18S and 28S rRNA signals measured by ethidium bromide staining. DNA for 32P-labelling reactions consisted of PCR amplification products of plasmids containing the coding sequence of the genes T1b, T2c and ICL1a. Each PCR reaction (100 mL) contained 50 ng of DNA matrix, 50 pmol each primer, 0.2 mm each dNTP and 2.5 U Taq DNA polymerase (Boehringer). The primers used for the amplification were: 5 0 -ATGTATAAATTAGCAGTCTGCACATTGC-3 0 (T1b sense), 5 0 -CAGCTCTTTGGAATTCAGC-3 0 (T1b antisense); 5 0 -ATGAAGACAATAATCCTTGCCTTAGCAC-3 0 (T 2c sense), 5 0 -TCAGATTTCTTCTCCAGCTGATTATCTTA-3 0 (T 2c antisense); 5 0 -GGCACGAAGAGGATAGT-3 0 (ICL1a sense), 5 0 -GCAAAGGTCTTTTTTGTCATAATG-3 0 (ICL1a antisense). All of the PCR reactions were carried out for one cycle of denaturation (1 min, 92 8C), and 30 cycles of denaturation (30 s, 92 8C), annealing (45 s, 54 8C) and extension (90 s, 72 8C) with a final extension at 728C for 10 min. Probes were synthesized by [a-32P]CTP 3000 Ci´mmol21 incorporation using a Random Primers Labelling System (Gibco-BRL), according to the supplier's protocol.

In vitro nuclear run-on transcription assays The protocol used is adapted from [36]. Chemicals were purchased from Sigma unless specified otherwise. Preparation of frameworks. One liter of cells (< 106 cells) was centrifuged, and the pellet (< 0.1 mL) was gently mixed with 0.9 mL ice cold extraction buffer (0.1 m sucrose, 0.1 m KCl, 2.5 mm MgCl2, 2.5 mm EGTA, 10 mm Hepes pH 6.5, 1% Triton X-100) and incubated on ice for 5 min. The permeabilized cells were washed once (centrifugation 1 min at 3000 g, 4 8C) in extraction buffer without Triton X-100. The resulting frameworks, consisting mainly of intact macronuclei and cortical cytoskeleton fragments, were resuspended in 250 mL transcription buffer I (0.05 mm Tris/HCl pH 8.1, 0.05 m KCl, 5 mm MgCl2, 1 mm spermine, 1 mm CaCl2, 2 mm dithiothreitol, 0.1 m sucrose, 25% glycerol, pH 8.1). 70-mL aliquots were frozen in liquid nitrogen and conserved at 280 8C until use. Transcription. Sixty microliters of frameworks were added to: 25 mL transcription buffer II (0.1 m Tris/HCl pH 8.1, 0.1 m KCl, 0.01 m MgCl2, 2 mm spermidine, 2 mm spermine, 4 mm putrescine, 6 mm dithiothreitol, 2 mm CaCl2, 1.2 mm aurintricarboxylic acid, pH 8.1), 2.5 mL ATP, CTP and GTP at 20 mm, 40 U RNasin (Boehringer) and 100 mCi [a-32P]UTP 3000 Ci´mmol21. This mix was incubated for 40 min at 25 8C. The RNA was extracted once with phenol/chloroform and once with chloroform, and the unincorporated nucleotides were removed with silica columns (QIAquick Nucleotide Removal

q FEBS 2000

Expression of Paramecium secretory protein genes (Eur. J. Biochem. 267) 3229

Kit; Qiagen). All of the labelled RNA, corresponding to 2.5±5.106 c.p.m., was used for one hybridization reaction. DNA blotting and hybridization conditions. Labelled transcripts were hybridized to DNA matrices prepared by PCR amplification of the coding region of T1b, T2c, T4a and b-tubulin genes. The primers used for the first two genes are described in `Preparation of radioactives probes for Northern blots'. For the other genes, the primers used are as follows: 5 0 -ATGGCTAGATCATTACAAATATTGGC-3 0 (T4a sense), 5 0 -TCAAAATACTTCTTCTCTGACTTGGAGG-3 0 (T4a antisense); 5 0 -ATGAGAGAAATCGTTCATATTCAAG-3 0 (b-tubulin sense), 5 0 -GTTGTGATAAAAATCACTTAGATTATC-3 0 (b-tubulin antisense).

For the b-tubulin gene probe, used as invariant control, genomic DNA was amplified, so that the probe contained sequences from the three nearly identical genes b PT1, b PT2 and b PT3 [37]. That b-tubulin transcription is not affected by exocytosis was supported by the constant values found in the experiments, and also by comparison with the transcription of a recently cloned histone H3 gene kindly provided by E. Meyer. The negative control was the pUC18 plasmid. 500 ng each DNA was mixed with 200 mL 5  NaCl/Cit, 0.4 m NaOH, and incubated for 30 min at 65 8C. The denatured DNA was then loaded on positively charged nylon membranes (Ambion) using a home-made dot-blot apparatus. The wells were flushed with the same buffer. After loading, the membrane was UV cross-linked and washed for 5 min in 2  NaCl/Cit.

Fig. 3. Gene activation in response to exocytosis. (A) Northern blot showing the simultaneous hybridization of total RNA with T1 and ICL1 probes. The experiment shows mRNA in wild-type cells before (0 min) and 20, 40, 60 and 120 min after AED stimulation and parallel mock-stimulation with buffer. (B) Normalized T1 mRNA levels at different times after AED-stimulated massive exocytosis, compared to the mock-stimulated control, obtained by quantification of the Northern blot shown in (A). The values represented are T1 mRNA normalized against ICL1 centrin mRNA. The moderate increase in T1 mRNA seen for the control mock-stimulated cells is the consequence of partial trichocyst release that occurs in the stimulation buffer, which contains 1 mm Ca21, as evaluated by the appearance of some flocculent material in the buffer and subnormal exocytotic capacity of the treated cells according to the picric acid test. The curves are representative of results obtained in two independent experiments. We verified that T2 and T4 subfamily mRNAs vary in the same way by stripping and reprobing the Northern blots. (C) Transcription of TMP genes measured by an in vitro nuclear run-on assay in wild-type cells in response to AED-stimulated massive exocytosis. Nuclear frameworks were isolated from cell populations at different times after exocytosis, the nascent transcripts were elongated in the presence of UT32P and hybridized with an array of gene sequences immobilized on a nylon membrane. The hybridization signals obtained for T1, T2 and T4 transcription initiation were normalized against the b-tubulin hybridization signal. (D) Comparison of transcription and mRNA levels in wild-type cells after exocytosis. The data for T1 mRNA levels after AED stimulation of wild-type cells was superimposed on the curve representing transcription initiation measured with the T2c probe in (B). (The values are similar to those measured with the T1b probe, but were chosen because of the stronger signal hence better signal to noise ratio.)

3230 A. Galvani and L. Sperling (Eur. J. Biochem. 267)

Prehybridizations were carried out at 40 8C for < 2 h, in 50% formamide (NorthernMax Prehybridization/hybridization bufferTM, Ambion). Hybridizations were carried out at the same temperature, in fresh buffer, for 48±72 h. Filters were finally washed several times at room temperature then at 50 8C, in 2  NaCl/Cit, 0.1% SDS. Exposure and quantification were as for Northern blots. RNA dot blots Total RNA was extracted from log-phase cultures as described above and a dilution series from 0.1 to 2.0 mg total RNA for each sample was applied to positively charged Nylon membranes (Ambion) according to the supplier's instructions, using a dot-blot apparatus. The DNA (ICL1a, b-tubulin, T1b) used to make 32P-labelled probes was prepared as described above and labelled by random priming. Hybridization conditions were the same as for the run-on transcription assay. The filters were successively hybridized with T1b, and b-tubulin probes and the signals quantified using a Phosphorimager (Molecular Dynamics). Quantification involved fitting each data set of eight dilution points by linear regression, the slopes providing the c.p.m.´mg21 total RNA. The linear regression coefficients R2 were all . 0.95 and most were . 0.98. The mRNA levels measured are presented as a multiple of the value found for wild-type cells in the same experiment.

q FEBS 2000

membrane. We measured the transcription of three TMP gene subfamilies (T1, T2 and T4). The b-tubulin gene was used as invariant control for normalization instead of ICL1a because the basic transcriptional level of the latter was too low to be detected reliably by this technique. The plasmid pUC18 served as a negative control. As shown in Fig. 3C, we observed rapid and strong transcriptional activation for the three subfamilies of TMP genes. This activation begins within 15 min of exocytosis, peaks around 30 min (maximal activation was 10- to 12-fold for each gene; n ˆ 3), decreases within 60 min to twice the basal level found in unstimulated cells, and does not return to the basal level during the 120-min period examined. This experiment also demonstrates clearly that TMP genes are co-ordinately regulated, as the time course and the degree of activation are the same for the three gene subfamilies under investigation. Nevertheless, the rate of transcription of these genes appears to be different (T2 . T4 . T1). This observation is difficult to interpret at present, as we do not know the exact number of genes in each subfamily that hybridize with the probes, and thus cannot be sure that the transcription rates are

R E S U LT S Increase of TMP mRNA levels after exocytosis in wild-type cells To determine whether TMP genes are activated after exocytosis, we studied the consequences of massive trichocyst discharge on TMP mRNA levels, using a centrin gene probe for internal normalization. The amount of steady-state T1 mRNA was measured at different times after secretagogue stimulation, by Northern blot analysis. As a control, part of the same wild-type culture was treated in parallel in exactly the same way, except that the stimulation buffer contained no secretagogue. Figure 3A,B shows that in wild-type cells treated with AED, T1 mRNA increases at least twofold after massive exocytosis. Most of the increase occurs between 20 and 40 min after secretagogue exposure, and the level of T1 mRNA remains elevated for at least 2 h after exocytosis. Doubling of the TMP mRNA pool is significant because T1 mRNA is abundant [32]. It thus appears that TMP gene expression is activated in response to exocytosis. The kinetic experiment shown in Fig. 3A,B was also carried out using lysozyme as secretagogue [38] and the same result was obtained (data not shown). Coordinate transcriptional activation of TMP genes To determine if the regulation observed is the result of transcriptional activation of the TMP genes rather than changes in mRNA stability, we performed nuclear run-on transcription assays. This technique allows assessment of the rate of transcription of several genes for a particular cellular state. Nuclear frameworks for in vitro transcription were prepared from populations of cells before and after stimulation with AED. Elongation of the nascent nuclear transcripts in the presence of UT32P provided a radioactive target that could be probed by hybridization with an array of genes fixed on a

Fig. 4. Steady-state mRNA levels after secretagogue stimulation of mutant cells. The values represented are T1 mRNA normalized against ICL1 centrin mRNA, measured by simultaneous hybridization of each Northern blot with a T1 and an ICL1 probe. In each experiment, the cell culture was divided in two and a mock-stimulated control was followed in parallel with the secretagogue-treated culture. Each set of curves is representative of results obtained in at least two independent experiments.

q FEBS 2000

Expression of Paramecium secretory protein genes (Eur. J. Biochem. 267) 3231

We performed nuclear run-on experiments with tam8, nd7-1 and nd9-3 mutant cells in order to look for variations that might not have been detected by Northern hybridization. Frameworks for in vitro transcription were prepared before and 32 min after AED treatment (Fig. 5). For each of the mutants, unlike the wild-type control which shows strong stimulation, the transcription of the TMP genes, normalized against b-tubulin transcription, is unaffected by AED treatment. Taken together, the Northern and the run-on data provide direct evidence that the calcium movements provoked by the secretagogue and necessary for exocytosis are not sufficient for activation of TMP gene expression. Fig. 5. TMP gene transcription in wild-type and mutant cells before and after stimulation with AED. For each cell line, the hybridization signals obtained by assaying cells before (±) and 32 min after (1) AEDtreatment are shown. The names of the genes fixed on the dot blot membrane are given on the right-hand side. In these experiments, the btubulin coding sequence served as invariant positive control and pUC18 plasmid DNA was the negative control. Only the wild-type cells show strong activation of TMP gene transcription after AED treatment.

really different because we may simply be detecting transcripts from different numbers of genes. The comparison of TMP gene expression measured by run-on transcription and of TMP mRNA measured by Northern blot in response to exocytosis, shown in Fig. 3D, suggests that, although the transcriptional activation is brief, it can probably account for the observed increase in mRNA. The peak of transcription coincides with the major increase in mRNA and this mRNA level stabilizes as transcription decreases to approximately twice its unstimulated rate. In other words, the mRNA curve represents the integral of the transcription curve, implying that TMP mRNA is quite stable.

The calcium movements associated with exocytosis are not sufficient for TMP gene activation To understand the origin of the signal(s) that leads to transcriptional activation of the TMP genes, we examined secretory mutants in which the reception and transduction of the signals for exocytosis can be dissociated from the exocytotic events themselves. We asked whether the calcium movements associated with exocytosis, in particular the calcium released from cortical stores which spills deep into the cytoplasm [22,23], can lead to the transcriptional activation according to the current implication of this second messenger in many instances of gene regulation (reviewed in [39,40]). We studied the mutants nd7-1 and nd9-3, defective for exocytotic membrane fusion although their trichocysts are attached to the cortical exocytotic sites, and tam8, whose trichocysts cannot attach to the cortex. It has been shown for these and other mutants with similar phenotypes that AED treatment leads, as in wild-type, to calcium influx [21] and to calcium release from internal stores [22,23]. The results of kinetic experiments using Northern blot analysis are shown in Fig. 4. We detected no changes in the amount of T1 mRNA in response to AED treatment in either nd7-1 or tam8 mutant cells. The curves corresponding to both stimulated and mock-stimulated cell populations are essentially invariant over the 2-h period following treatment, compared to the curves obtained for wild-type cells (see Fig. 3A and B).

The state of occupation of the exocytotic sites is correlated with TMP gene expression Quantification of the run-on hybridization signals shown in Fig. 5 indicates that the basal transcription level in the two mutants with docked trichocysts, nd7-1 and nd9-3, is comparable to the level found in unstimulated wild-type cells. Interestingly, the value measured for the mutant with undocked trichocysts, tam8 (n ˆ 3), is twofold higher (P , 0.001) than that measured for either unstimulated wild-type or nd mutant cells (n ˆ 4). This prompted us to examine TMP mRNA levels of different mutants with docked or undocked trichocysts, using quantitative RNA dot blot analysis. TMP mRNA levels for a number of secretory mutants in their resting, i.e. unstimulated state were measured, as shown in Fig. 6. The phenotypes of the mutants and the values we measured are presented in Table 2. In so far as possible, we have studied mutants in which the site of action of the mutations had previously been determined by rescue experiments, as given in Table 1. We found that mutants with docked trichocysts, nd3-5 and nd6-1 had T1 mRNA levels similar to that of unstimulated wild-type cells. Mutants with undocked trichocysts, tam8 and tam54, presented values approximately 2.5-fold higher. Two leaky alleles of the TAM6 gene, tam6-1 and tam6-2, which confer intermediate phenotypes, have intermediate mRNA levels. The tam6-1 cells present a subnormal exocytotic phenotype and a T1 mRNA level only slightly higher than that of wild-type cells or of nd3-5 and nd6-1 mutant cells. Table 2. RNA dot blot determination of T1 mRNA levels in secretory mutants. The results of the quantification of the dot blots shown in Fig. 6 are given along with the phenotypes of the different strains. The state of occupation of the exocytotic sites is 1 or ± in each case with the exception of the two leaky tam6 mutants; (1) indicates subnormal exocytotic capacity while :1 means that approximately 50 trichocysts per cell are docked and secretable. The values for T1 mRNA (bold type) have been normalized against b-tubulin mRNA and scaled with respect to the wild-type value. The numbers in parentheses represent the linear regression coefficients found respectively for the T1 and the b-tubulin dilution series.

Strain

Site occupation

T1 mRNA (R2 coefficient)

Wild-type tam54 nd3-5 nd6-1 tam6-1 tam6-2 tam8

1 ± 1 1 (1) :1 ±

1.0 2.6 1.1 1.2 1.4 1.7 2.6

(0Š.97,0.98) (0Š.99,0.98) (0Š.99,0.99) (0Š.96,0.99) (0Š.98,0.99) (0Š.99,0.99) (0Š.99,0.98)

3232 A. Galvani and L. Sperling (Eur. J. Biochem. 267)

Fig. 6. Comparison of T1 mRNA levels in different secretory mutants using quantitative RNA dot blots. The same dot blots, representing dilution series for each sample of 0.1, 0.2, 0.3, 0.5, 0.7, 1.0, 1.5 and 2.0 mg of total RNA, were hybridized with a T1 probe, stripped, and hybridized with a b-tubulin probe.

tam6-2 cells contain < 50 docked and secretable trichocysts judging by the picric acid test and phase contrast microscopy, and these cells present a T1 mRNA level higher than that of tam6-1 cells but lower than that found for tam8 and tam54 cells in which none of the docking sites are occupied. Several other mutants were examined by side-to-side comparison on Northern blots (data not shown). We found that nd7-1, like nd3-5 and nd6-1, had a T1 mRNA level comparable to that of unstimulated wild-type cells, while the mutants with undocked trichocysts tam1, tam8, tam10, tam11 and tam54 all presented similar values, around 2.5-fold higher. Thus the correlation between unoccupied docking sites and elevated mRNA levels is strictly observed, whatever the site of action of the different mutations. This inverse correlation between docking site occupation and T1 mRNA is further strengthened by a study of gene silencing [41]. Inactivation of either T1 or T4 genes by gene silencing yields cells with aberrantly shaped trichocysts unable to attach to the cortical docking sites which are thus uniformly unoccupied. Ruiz et al. [41] found that, although mRNA for the silenced gene subfamily (e.g. T1 mRNA) is specifically reduced, mRNA for other TMP gene subfamilies (e.g. T4 mRNA) is increased at least twofold, as in the Mendelian mutants with undocked trichocysts.

DISCUSSION We have used Paramecium as a model to study the effects of regulated exocytosis on the expression of genes encoding secretory proteins. We found that the stimulation of massive exocytosis leads to transient, 10-fold transcriptional activation of the TMP genes encoding the secretory proteins, resulting in a 2.5-fold increase in the pool of TMP mRNA. The kinetics of the activation are consistent with the wave of biosynthetic activity following exocytosis characterized by a previous morphological study [42] and indicate that Paramecium can react to intense secretory activity by stimulating the machinery necessary to make new secretory granules. This response appears to be a conserved feature of the regulated secretory pathway from protozoa to professional secretory cells in metazoa. We used a collection of secretory mutants to see whether the signals for stimulus/secretion coupling are also involved in the transcription activation. Secretagogue application in Paramecium leads to an increase in intracellular free calcium in the cortical region [22]. In one mutant in which a secretagogue-induced calcium influx is specifically blocked, no exocytosis occurs [21], providing a direct argument that calcium ions are necessary for regulated membrane fusion in Paramecium as in other systems [43]. We performed

q FEBS 2000

experiments with secretory mutants that display the same calcium movements in response to secretagogue as wild-type cells, even though the supposed downstream targets of the Ca21 are defective or absent. Most significantly, for nd9-1 mutant cells at the nonpermissive temperature, Klauke and Plattner [22] used calcium imaging techniques with high spatial and time resolution to show that both the initial increase in [Ca21]i at the site of AED application and the spill-over into deeper cell regions were the same as for wild-type cells. In one mutant with no docked trichocysts (trichless [22]), the calcium reaches the nuclear region more rapidly and the signal has greater amplitude. We found that these secretagogue-induced modulations in the amplitude and localization of intracellular free calcium ions, which clearly reach the nuclear region in the different types of mutants studied here, are not sufficient to activate TMP gene expression (although the increase in cytosolic free calcium may well turn out to be necessary for the response, once something else has activated it). This result is surprising, as calcium ions are involved in a great variety of signalling events [39,40,44,45]. Moreover, calcium ions have been shown to activate both secretion of chromaffin granules and transcription of chromogranin A in pheochromocytoma cells, although achievement of the transcriptional activation depends upon the precise route of entry of calcium into the cytosol [46]. However, our results provide evidence that the state of occupation of trichocyst docking sites at the cell cortex affects secretory protein gene expression. We found a strict correlation between TMP mRNA levels and site occupation; cells with empty sites present 2.5-fold more TMP mRNA than cells in which the sites are occupied by docked trichocysts. We consider that our data go beyond simple correlation as might arise from pleiotropic mutations affecting the same structure or enzymatic pathway. Indeed, the correlation is sustained in three experimental situations: (a) Mendelian mutations in genes with dictinct sites of action i.e. trichocyst, cytosol or plasma membrane; (b) cells with altered phenotypes created by homology-dependent silencing of the secretory protein genes themselves; and (c) wild-type cells in which the sites are empty following exocytosis of the trichocysts. In the Mendelian mutants, and in wild-type cells after exocytosis, we have shown that the increase in TMP mRNA is acheived by transcriptional activation. After exocytosis in wild-type cells, strong but transient activation of TMP genes results in an increase of the TMP mRNA pool by a factor of 2.5 within 45 min. As soon as the mRNA pool reaches this level, transcription decreases to twice the basal rate. Transcription only returns to the basal rate in cells with a full complement of docked trichocysts. In the Mendelian mutants in which the exocytotic sites are empty at all times, transcription at twice the basal rate is sufficient to maintain a 2.5-fold increase in TMP mRNA. Taking into account the ensemble of the data, we propose that the state of the docking sites plays a role in stimulus/ transcription coupling in Paramecium. The exocytotic sites might even be responsible for all of the activation, if we add an ad hoc hypothesis, namely that there is a negative feedback loop resembling the translation-dependent negative feedback mechanism found in the serum response of immediate early genes [47]. Such negative feedback, as a function of the amount of mRNA being translated, could account for the inactivation of the response down to twice basal level starting 30 min after exocytosis. Alternatively, it is possible that additional signals converge onto TMP promoters, activated by the final exocytotic events (membrane fusion or released contents). Whether the

q FEBS 2000

Expression of Paramecium secretory protein genes (Eur. J. Biochem. 267) 3233

exocytotic sites account for part or all of the response to exocytosis, we suggest that in their unoccupied configuration they activate a signalling pathway to the nucleus that enhances TMP gene transcription. The exocytotic sites are already known to communicate with the nucleus, via the cytoskeleton. The tam mutants, in addition to their exocytosis-deficient phenotype, are characterized by abnormal shape, position and division of the somatic macronucleus [51]. These nuclear anomalies are the result of dramatic perturbation of the cytoskeleton, as revealed by immunocytochemical examination of the microtubule networks [48]. Cytoplasmic microtubules, which in wild-type cells run perpendicular to the cortex where they are nucleated to the macronucleus where many of them are tethered, lie parallel to the cortex in mutants with no docked trichocysts to constrain their path. Most significantly, the same configuration can be seen following massive exocytosis, while phenocopies of the characteristic nuclear defects of the tam mutants can be obtained by depolymerization of the internal microtubule network with nocodazole [48]. That the configuration of the cytoplasmic microtubule array could be involved in relaying information to the nucleus for activation of TMP genes seems possible; microtubule depolymerization is known to activate at least one well studied transcription factor, NF-kB [49], while the microtubule stabilizing drug taxol blocks NF-kB activation [50]. Granule docking in metazoan secretory cells involves reorganization of the cortical actin cytoskeleton [2], so that implication of the cytoskeleton in regulation of genes required for granule biogenesis could be of general relevance.

ACKNOWLEDGEMENTS We thank M.-C. Gautier, whose preliminary data prompted us to undertake the present study. We thank J. Beisson and J. Cohen for encouragement throughout the project and many useful comments on the manuscript. We thank P. Dupuis-Williams and L. Vayssie for critical reading of the manuscript. A. G. is supported by a graduate fellowship from the MinisteÁre de l'Education Nationale, de la Recherche et de la Technologie (MENRT). The project was financed by ACC-SV6 contract no. 9506004 (MENRT) and by the Microbiology Program of the MENRT (Programme de Recherche Fondamentale en Microbiologie et Maladies Infectieuses et Parasitaires).

REFERENCES 1. Arvan, P. & Castle, D. (1998) Sorting and storage during secretory granule biogenesis: looking backward and looking forward. Biochem. J. 332, 593±610. 2. Martin, T. (1997) Stages of regulated exocytosis. Trends Cell Biol. 7, 271±276. 3. Haddad, A. & Turkewitz, A.P. (1997) Analysis of exocytosis mutants indicates close coupling between regulated secretion and transcription activation in Tetrahymena. Proc. Natl Acad. Sci. USA 94, 10675±10680. 4. Holthuis, J.C., Jansen, E.J., van Riel, M.C. & Martens, G.J. (1995) Molecular probing of the secretory pathway in peptide hormoneproducing cells. J. Cell Sci. 108, 3295±3305. 5. Eiden, L.E., Giraud, P., Dave, J.R., Hotchkiss, A.J. & Affolter, H.U. (1984) Nicotinic receptor stimulation activates enkephalin release and biosynthesis in adrenal chromaffin cells. Nature 312, 661±663. 6. Oyarce, A.M., Hand, T.A., Mains, R.E. & Eipper, B.A. (1996) Dopaminergic regulation of secretory granule-associated proteins in rat intermediate pituitary. J. Neurochem. 67, 229±241. 7. Leibiger, I.B., Leibiger, B., Moede, T. & Berggren, P.O. (1998) Exocytosis of insulin promotes insulin gene transcription via the insulin receptor/PI-3 kinase/p70 s6 kinase and CaM kinase pathways. Mol. Cell 1, 933±938. 8. Schuppin, G.T. & Rhodes, C.J. (1996) Specific co-ordinated regulation

9.

10. 11.

12.

13.

14.

15. 16.

17.

18.

19.

20.

21. 22.

23.

24.

25.

26. 27. 28.

29.

of PC3 and PC2 gene expression with that of preproinsulin in insulin-producing beta TC3 cells. Biochem. J. 313, 259±268. Tang, K., Wu, H., Mahata, S.K., Taupenot, L., Rozansky, D.J., Parmer, R.J. & O'Connor, D.T. (1996) Stimulus-transcription coupling in pheochromocytoma cells. Promoter region-specific activation of chromogranin a biosynthesis. J. Biol. Chem. 271, 28382±28390. VayssieÂ, L., Skouri, F., Sperling, L. & Cohen, J. (2000) Molecular genetics of regulated secretion in Paramecium. Biochimie 82, 1±20. Knoll, G., Braun, C. & Plattner, H. (1991) Quenched flow analysis of exocytosis in Paramecium cells: time course, changes in membrane structure, and calcium requirements revealed after rapid mixing and rapid freezing of intact cells. J. Cell Biol. 113, 1295±1304. Beisson, J., Lefort-Tran, M., Pouphile, M., Rossignol, M. & Satir, B. (1976) Genetic analysis of membrane differentiation in Paramecium: freeze-fracture study of the trichocyst cycle in wild type and mutant strains. J. Cell Biol. 69, 126±143. Lefort-Tran, M.K., Aufderheide, M., Pouphile, M., Rossignol, M. & Beisson, J. (1981) Control of exocytotic processes: cytological and physiological studies of trichocyst mutants in Paramecium tetraurelia. J. Cell Biol. 88, 301±311. Pouphile, M., Lefort-Tran, M., Plattner, H., Rossignol, M. & Beisson, J. (1986) Genetic dissection of the morphogenesis of exocytosis sites in Paramecium. Biol. Cell 56, 151±162. Janisch, R. (1972) Pellicle of Paramecium caudatum as revealed by freeze-etching. J. Protozool. 19, 470±472. Plattner, H., Miller, F. & Bachmann, L. (1973) Membrane specializations in the form of regular membrane-to-membrane attachment sites in Paramecium. A correlated freeze-fracture and ultrathin sectioning analysis. J. Cell Sci. 13, 687±719. Hausmann, K. & Allen, R.D. (1976) Membrane behavior of exocytic vesicles. II. Fate of the trichocyst membranes in Paramecium after induced trichocyst discharge. J. Cell Biol. 69, 313±326. Pape, R. & Plattner, H. (1985) Synchronous exocytosis in Paramecium cells. V. Ultrastructural adaptation phenomena during re-insertion of secretory organelles. Eur. J. Cell Biol. 36, 38±47. Skouri, F. & Cohen, J. (1997) Genetic approach to regulated exocytosis using functional complementation in Paramecium: identification of the ND7 gene required for membrane fusion. Mol. Biol. Cell 8, 1063±1071. Kerboeuf, D., Le Berre, A., Dedieu, J.C. & Cohen, J. (1993) Calmodulin is essential for assembling links necessary for exocytotic membrane fusion in Paramecium. EMBO J. 12, 3385±3390. Kerboeuf, D. & Cohen, J. (1990) A Ca21 influx associated with exocytosis is specifically abolished in a Paramecium exocytotic mutant. J. Cell Biol. 111, 2527±2535. Klauke, N. & Plattner, H. (1997) Imaging of Ca21 transients induced in Paramecium cells by a polyamine secretagogue. J. Cell Sci. 110, 975±983. Erxleben, C., Klauke, N., Flotenmeyer, M., Blanchard, M.P., Braun, C. & Plattner, H. (1997) Microdomain Ca21 activation during exocytosis in Paramecium cells. Superposition of local subplasmalemmal calcium store activation by local Ca21 influx. J. Cell Biol. 136, 597±607. Stelly, N., Halpern, S., Nicolas, G., Fragu, P. & Adoutte, A. (1995) Direct visualization of a vast cortical calcium compartment in Paramecium by secondary ion mass spectrometry (SIMS) microscopy: possible involvement in exocytosis. J. Cell Sci. 108, 1895±1909. Stelly, N., Mauger, J.P., Claret, M. & Adoutte, A. (1991) Cortical alveoli of Paramecium: a vast submembranous calcium storage compartment. J. Cell Biol. 113, 103±112. Sonneborn, T.M. (1974) Paramecium aurelia. In Handbook of Genetics (King, R., ed.), pp. 469±594. Plenum Publishing Corp, New York. Sonneborn, T.M. (1970) Methods in Paramecium research. Methods Cell Physiol. 4, 241±339. Plattner, H., Matt, H., Kersken, H., Haacke, B. & StuÈrzl, R. (1984) Synchronous exocytosis in Paramecium cells. A. A novel approach. Exp. Cell Res. 151, 6±13. Plattner, H., Sturzl, R. & Matt, H. (1985) Synchronous exocytosis in

3234 A. Galvani and L. Sperling (Eur. J. Biochem. 267)

30.

31.

32.

33.

34.

35.

36.

37. 38.

39. 40. 41.

42.

Paramecium cells. IV. Polyamino compounds as potent trigger agents for repeatable trigger-redocking cycles. Eur. J. Cell Biol. 36, 32±37. Pollack, S. (1974) Mutations affecting the trichocysts in Paramecium aurelia. I. Morphology and description of the mutants. J. Protozool. 21, 352±362. Chomczynski, P. & Sacchi, N. (1987) Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal. Biochem. 162, 156±159. Madeddu, L., Gautier, M.-C., VayssieÂ, L., Houari, A. & Sperling, L. (1995) A large multigene family codes for the polypeptides of the crystalline trichocyst matrix in Paramecium. Mol. Biol. Cell 6, 649±659. Gautier, M.-C., Sperling, L. & Madeddu, L. (1996) Cloning and sequence analysis of genes coding for Paramecium secretory granule (trichocyst) proteins. A unique protein fold for a family ofpolypeptides with different primary structures. J. Biol. Chem. 271, 10247±10255. Madeddu, L., Klotz, C., Le Caer, J.-P. & Beisson, J. (1996) Characterization of centrin genes in Paramecium. Eur. J. Biochem. 238, 121±128. VayssieÂ, L., Sperling, L. & Madeddu, L. (1997) Characterization of multigene families in the micronuclear genome of Paramecium tetraurelia reveals a germline specific sequence in an intron of a centrin gene. Nucleic Acids Res. 25, 1036±1041. Gilley, D., Rudman, B.M., Preer, J.R. Jr & Polisky, B. (1990) Multilevel regulation of surface antigen gene expression in Paramecium tetraurelia. Mol. Cell Biol. 10, 1538±1544. Dupuis, P. (1992) The b-tubulin genes of Paramecium are interrupted by two 27 bp introns. EMBO J. 11, 3713±3719. Hennessey, T.M., Kim, M.Y. & Satir, B.H. (1995) Lysozyme acts as a chemorepellent and secretagogue in Paramecium by activating a novel receptor-operated Ca21 conductance. J. Membr. Biol. 148, 13±25. Finkbeiner, S. & Greenberg, M.E. (1997) Spatial features of calciumregulated gene expression. Bioessays 19, 657±660. Berridge, M.J., Bootman, M.D. & Lipp, P. (1998) Calcium ± a life and death signal. Nature 395, 645±648. Ruiz, F., VayssieÂ, L., Klotz, C., Sperling, L. & Madeddu, L. (1998) Homology-dependent gene silencing in Paramecium. Mol. Biol. Cell. 9, 931±943. Garreau de Loubresse, N. (1993) Early steps of the secretory pathway in Paramecium: ultrastructural, immunocytochemical and genetic analysis of trichocyst biogenesis. In Membrane Traffic in Protozoa (Plattner, H., ed.), pp. 27±59. JAI Press Inc, Greenwich, CT, USA.

q FEBS 2000 43. Burgoyne, R.D. & Morgan, A. (1993) Regulated exocytosis. Biochem. J. 293, 305±316. 44. Bading, H., Ginty, D.D. & Greenberg, M.E. (1993) Regulation of gene expression in hippocampal neurons by distinct calcium signaling pathways. Science. 260, 181±186. 45. Hardingham, G.E., Cruzalegui, F.H., Chawla, S. & Bading, H. (1998) Mechanisms controlling gene expression by nuclear calcium signals. Cell Calcium 23, 131±134. 46. Tang, K., Wu, H., Mahata, S.K., Mahata, M., Gill, B.M., Parmer, R.J. & O'Connor, D.T. (1997) Stimulus coupling to transcription versus secretion in pheochromocytoma cells. Convergent and divergent signal transduction pathways and the crucial roles for route of cytosolic calcium entry and protein kinase C. J. Clin. Invest. 100, 1180±1192. 47. Greenberg, M.E., Hermanowski, A.L. & Ziff, E.B. (1986) Effect of protein synthesis inhibitors on growth factor activation of c- fos, c-myc, and actin gene transcription. Mol. Cell Biol. 6, 1050±1057. 48. Cohen, J. & Beisson, J. (1988) The Cytoskeleton. In Paramecium (GoÈrtz, H.-D., ed.), pp. 363±391. Springer-Verlag, Berlin, Heidelberg. 49. Rosette, C. & Karin, M. (1995) Cytoskeletal control of gene expression: depolymerization of microtubules activates NF-kappa B. J. Cell Biol. 128, 1111±1119. 50. Spencer, W., Kwon, H., Crepieux, P., Leclerc, N., Lin, R. & Hiscott, J. (1999) Taxol selectively blocks microtubule dependent NF-kB activation by phorbol ester via inhibition of IkBa phosphorylation and degradation. Oncogene. 18, 495±505. 51. Cohen, J. & Beisson, J. (1980) Genetic analysis of the relationships between the cell surface and the nuclei in Paramecium tetraurella. Genetics 95, 797±818. 52. Beisson, J., Cohen, J., Lefort-Tran, M., Pouphile, M. & Rossignol, M. (1980) Control of membrane fusion in exocytosis. Physiological studies on a Paramecium mutant blocked in the final step of the trichocyst extrusion process. J. Cell Biol. 85, 213±227. 53. Beisson, J. & Rossignol, M. (1975) Movements and positionning of organelles in Paramecium aurelia. In Molecular Biology of Nucleocytoplasmic Relationships (Puiseux-Dao, S., ed.), pp. 291±294. Elsevier, Amsterdam. 54. Aufderheide, K.J. (1978) Genetic aspects of intracellular motility: cortical localization and insertion of trichocysts in Paramecium tetraurelia. J. Cell Sci. 31, 259±273. 55. Allen, S.L., Altschuler, M.I., Bruns, P.J., Cohen, J., Doerder, F.P., Gaertig, J., Gorovsky, M., Orias, E. & Turkewitz, A. (1998) Proposed genetic nomenclature rules for Tetrahymena thermophila, Paramecium primaurelia and Paramecium tetraurelia. The Seventh International Meeting on Ciliate Molecular Biology Genetics Nomenclature. Genetics 149, 459±462.