DEVELOPMENTAL GENETICS 13:167-173 (1992)
Genetic Characterization of the Secretory Mutant MS-1 of Tetrahymena thermophila: Vacuolarization and Block in Secretion of Lysosomal Hydrolases Are Caused by a Single Gene Mutation PETER HUNSELER AND ARNO TIEDTKE Institute of General Zoology and Genetics, University of Munster, Munster, Federal Republic of Germany
ABSTRACT The genetics and phenotypic features (light and electron microscopy) of a secretory mutant, MS-1, of Tetrahyrnena therrnophila blocked in secretion of lysosomal acid hydrolases have been analyzed. Although blocked constitutively in secretion, MS-1 contains active lysosomal hydrolases in amounts equivalent to the wild type. The 3:l segregation in F-2 in sib crosses and the 1 : l segregation in test crosses indicate that the block in secretion of lysosomal hydrolases is controlled by a recessive single gene locus termed sec. The sec allele of MS-1 proved also to be responsible for the highly vacuolarized phenotype the mutant developed when it was transferred from nutrient medium into buffers of low ionic strength. Deletion mapping by crossing MS-1with nullisomic strains, all secreting lysosomal hydrolases at wildtype rates, was performed. The sec phenotype was expressed in monosomic-4 progeny only, indicating that the sec allele is located on chromosome 4 of T. therrnophila. o 1992 WiIey-Liss, Inc. Key words: Acid hydrolases, sec allele, phenotype
INTRODUCTION The ciliated protozoon Tetrahymena thermophila secretes constitutively large amounts of acid hydrolases into the culture medium [Muller, 1972; Blum and Rothstein, 1975; Tiedtke and Rasmussen, 19891. Growing and multiplying cells replace the secreted hydrolases constantly, while cells suspended in starvation medium cannot replace the secreted enzymes and may lose up to 80% of the enzyme activities initially present in the cells. We recently established a screening procedure for isolation of mutants with blocks in secretion of lysosomal enzymes in order to get tools for a genetic dissection of the pathway these secreted enzymes take through the cell. The secretory mutant MS-1 dealt with
0 1992 WILEY-LISS, INC.
in this study is constitutively blocked in secretion of lysosomal acid hydrolases. It contains intracellular amounts of hydrolases similar to those in wild-type cells [Hunseler et al., 19871. Studies on the biosynthesis of two lysosomal hydrolases, P-hexosaminidase [Hunseler et al., 19881and a-glucosidase (unpublished data), have shown no detectable differences in synthesis and processing of these enzymes in mutant and wild-type cells. The mutant grows on a diet of bacteria as well as do wild-type cells, indicating that it contains intracellularly fully functional lysosomes. We therefore assume that a late step in the secretory pathway of lysosomal acid hydrolases is affected by the mutation [Tiedtke et al., 1988al. The sec mutant MS-1 has already become a valuable tool in studies on the biological significance of extracellular acid hydrolases [Tiedtke and Rasmussen, 1988; Florin-Christensen et al., 19901 and on the control of release of acid hydrolases [Tiedtke et al., 1988bl. Studies on sec and ex0 mutants (the latter blocked in secretion of mucocysts [Orias et al., 19831revealed that Tetrahymena possesses at least two independent secretory pathways: the regulated exocytosis of mucocysts and the constitutive secretion of acid hydrolases. While a n ex0 mutant (SB 2811, possessing no mature mucocysts, released acid hydrolases to the same extent as wild-type cells, MS-1 nonetheless is able to extrude mucocysts [Tiedtke et al., 1988a, and unpublished observations]. Upon starvation in inorganic buffers of low ionic strength, MS-1 develops numerous translucent vacuoles, which progressively enlarge and finally fuse to
Received for publication October 7, 1991; accepted November 11, 1991. Address reprint requests to Dr. Arno Tiedtke, Institute for General Zoology and Genetics, Schlopplatz 5, D-4400Miinster, Federal Republic of Germany.
168
HUNSELER AND TIEDTKE -
Strain cu 399 CU 428 cu 353 CU 361
cu 383
TABLE 1. Tetrahyrnena Strains Micronuclear genotype ChxlChx MprlMpr Nulli-5; MprlMpr Nulli-3; ChxlChx; MprlMpr Nulli-4; ChxlChx; MprIMpr
form a huge vacuole that dramatically changes the morphology of the cells. In the present study, we characterized genetically the expression and transmission of the sec allele and show that the block in secretion of lysosomal hydrolases and the change in cell shape are caused by the same single gene mutation.
Macronuclear phenotype Cycloheximide sensitive 6-Methylpurine sensitive 6-Methylpurine sensitive Cycloheximide sensitive; 6-Methylpurine sensitive Cycloheximide sensitive; 6-Methylpurine sensitive
Mating type
IV VII IV IV
VI
Drugs Drug resistance of progeny was tested in PPYS/PS medium containing 6-methylpurine and/or cycloheximide in final concentrations of 15 pg X m1-I. Crosses Standard crosses were carried out under axenic conditions in Dryl’s solution principally as described by Orias and Bruns [1976]. Pairs were isolated with micropipettes into drops of PPYS/PS medium on Petri dishes and allowed to grow at 30°C for 48 hr. Synclones were directly replicated into microtiter plates, containing 0.1 ml medium per well.
MATERIALS AND METHODS Strains The secretory mutant MS-1 of T. thermophila, blocked constitutively in secretion of lysosomal hydrolases, was obtained from progeny of clone CU 399, previously subjected to mutagenesis with 10 pg x ml-’ N-methyl-N’-nitro-N-nitrosoguanidine, mated with Identification of True Progeny clone C* followed by short-circuit genomic exclusion with positive selection for mating [Bruns and BrusTrue progeny are clones that possess a newly develsard, 19741. The genetic marker being used for the se- oped macronucleus after conjugation. In many cases, at lection for mating was a dominant allele (Chx)confer- least one of the parental clones was a functional hetring resistance to cycloheximide [Bleyman and Bruns, erokaryon, phenotypically sensitive to cycloheximide 19771. Owing to the nature of short-circuit genomic or 6-methylpurine. True progeny clones therefore could exclusion, the secretory mutant may be presumed to be be identified by their newly acquired resistance to homozygous for the Chx allele as well as for the allele these drugs. In rare cases when these markers were conferring the block in secretion of lysosomal hydro- absent, true progeny clones were identified by their lases. The screening for mutants blocked in secretion of sexual immaturity, a s indicated by a mating test (malysosomal hydrolases has been reported elsewhere turity test) as described below. [Hunseler et al., 19871. Clones that did not conjugate retain the old macroThe following stocks were used in the genetic anal- nucleus and remain sexually mature. Clones were ysis of MS-1 (Table 1).Following the nomenclature pro- tested against at least two nonparental mating types posed by Bruns and Brussard [1974], CU 383, for ex- and were considered immature and were retained for ample, has a micronuclear genotype homozygous further work if they did not form pairs. Immature resistant to cycloheximide and 6-methylpurine; in ad- clones were transferred until maturity was reached dition, chromosome 4 is missing in the micronucleus. and accepted as true progeny by the most stringent In the macronuclear phenotype, the strain is sensitive criterion of having nonparental mating types. to both drugs and expresses mating type VI. A collection of mating type tester strains (mating types I-VII) Mass Selection of Progeny and the functional heterokaryons were donated by Dr. In crosses with nullisomic strains, small samples (10 P. Bruns (Cornell University, Ithaca, NY). p1) of the mating mixture were distributed 12 h r after mixing of the cells to wells of microtiter plates containMedia ing 90 p1 of PPYS/PS medium per well. Cycloheximide Stock cultures were maintained in test tubes on 1% and 6-methylpurine dissolved in the medium at 30 pg proteose peptone (Difco). Mass cultures were grown a t ml-’ each were added in equivalent volumes to the 30°C on 1% proteose peptone supplemented with 0.1% wells. Nonconjugant clones, being sensitive to the yeast extract (Difco) 0.003% sequestrene (Geigy) and drugs, are killed by 3 days in the presence of the drugs. 250 pg x ml-’ each of penicillin G and streptomycin Surviving cells were then removed from the wells and cloned. The clones were grown in the presence of both sulphate (Sigma) (abbreviated PPYS/PS medium).
GENETICS OF THE SECRETORY MUTANT MS-1
169
drugs in final concentrations of 15 pg x ml-' each and were tested for vacuolarization and secretion of lysosoma1 hydrolases.
uranyl acetate and examined with a Siemens electron microscope.
Mating Tests Mating tests were performed immediately after conjugation to identify potential true exconjugant clones. Exconjugant clones and strains of mating type testers were coreplicated into V-bottom microtiter plates, grown overnight a t 30°C, and washed into Dryl's solution so that pair formation could take place in the same well. Failure of pair formation with any of the tester strains in four independent experiments was taken as indication of immature true exconjugant clones. Mating tests were repeated after exconjugant clones had matured to determine the mating type of the clone by testing it against all mating type testers.
RESULTS Morphological Characteristics of the sec Mutant MS 1 MS-1 is blocked in secretion of lysosomal acid hydrolases and releases less than 5% of its total enzyme activities into the extracellular fluids [Hunseler et al., 19871. When starved in Dryl's solution or other lowmolarity buffers for 4-6 h r to initiate conjugation, the mutant exhibits a dramatic change in its morphology (Figs. l a and 2): translucent vacuoles fuse and enlarge progressively until the cells contain one or two huge vacuoles. Wild-type cells starved for the same time become slim and contain only a few small vacuoles (Fig. lb). Unlike the mutant, wild-type cells had released by the same time 80% of their initial amount of @-hexosaminidase and 40-50% of acid phosphatase activities to the starvation medium. With the breeding analysis of the see allele we therefore also scored for the vacuolarized phenotype of progeny clones to test whether the sec- and the vacuolarized phenotypes are coexpressed or independently transferred to progeny clones.
Test for Secretion of Lysosomal Hydrolases and Vacuolarized Progeny Five milliliter samples of progeny clones were grown to cell concentrations of 5 x lo5 cells x ml-', washed twice with and resuspended in 5 ml of Dryl's solution, and kept for 4 h r a t 30°C. The appearance of vacuolarized progeny clones was scored. The extracellular and total activities of lysosomal hydrolases were determined a s follows; one-half of a sample was layered on a cushion of 10% Ficoll (Pharmacia) and centrifuged for 3 min a t 600 g. The supernatant was removed and assayed for the extracellular (secreted) lysosomal enzyme activities. The other one-half of the sample was frozen and thawed and then sonicated on ice for 30 sec a t 150 mA (Bandelin Sonorex). Particulate materials were sedimented at 10,000 g for 4 min (Runne model 85-1 microcentrifuge). The clarified supernatant was used to measure the total activities of lysosomal enzymes, i.e., cellular and extracellular activities, and to determine the protein concentration by the method of Lowry et al. [19511. The activities of N-acetyl-P-D-hexosaminidase (EC 3.2.1.52 = 6-hexosaminidase) and acid phosphatase (EC 3.1.3.2) were determined as described by Tiedtke [1983]. One unit was defined a s the hydrolase activity that releases 1 pmol p-nitrophenollmin at 37°C. Progeny clones were regarded a s blocked in secretion of lysosomal hydrolases if the secreted activities of both enzymes were less than 10% of those found in the total system. Light and Electron Microscopy (EM) Cells were washed twice with 50 mM PIPES buffer, pH 7.2; resuspended in a small volume of the buffer; and fixed with glutaraldehyde at a final concentration of 2% for 30 min at room temperature. Light micrographs were taken with phase optics using a Zeiss photomicroscope. For EM, the cells were postfixed with 1.5% OsO, for 30 min a t room temperature, washed, dehydrated in a n ethanol series, and embedded in Epon. Thin sections were stained with lead citrate and
Genetic Analysis of the Mutant MS-1 Blocked in Secretion of Acid Hydrolases Analysis of F, progeny. Due to the nature of shortcircuit genomic exclusion, the cycloheximide-resistant mutant MS-1 was presumed to be homozygous for the Chx allele and, as a whole-genome homozygous ( = isozygous) clone, necessarily also homozygous for the allele(s), conferring the block in secretion of acid hydrolases and the expression of the vacuolarized phenotype. To test this assumption, we crossed MS-1 (previously shown to express mating type 11) to CU 428. From this cross, conjugant pairs were isolated and the synclones tested for sexual immaturity. From sexually immature synclones, single cells were cloned and tested for resistance against both drugs, cycloheximide and 6-methylpurine. Only nine of 192 viable synclones fulfilled these criteria and were regarded as true F, exconjugants. The nine clones secreted acid hydrolases in amounts similar to their parent strain, CU 399, and none of them developed the vacuolarized phenotype of MS-1. The allele(s) conferring the block in secretion of acid hydrolases and expression of the vacuolarized phenotype behaved recessively (Table 2). The genetic constitution of MS-1 therefore was assumed to be seclsec, ChxlChx (cycloheximide resistant 11). Phenotypic assortment of F, progeny. The nine heterozygous F, clones were subcloned daily and tested for sexual maturity. Two clones with mating types different from both parent strains, 9a and l l a , were chosen for further analysis. In the course of maturity tests,
170
HUNSELER AND TIEDTKE
Fig. 1. a,b Phase-contrast micrographs of cells starved for 3 h r in Dryl’s solution. Highly translucent vacuoles are observed only in the sec mutant MS-1(a); the wild type contains only a few small vacuoles (b). Bar = 10 p m
a few cells were observed in these clones showing the vacuolarized phenotype of MS-1. These cells were subcloned until they uniformly expressed the vacuolarized phenotype. When these assorters, 9a ass. and l l a ass., were tested for secretion of acid hydrolases, they were found to be unable to secrete the enzymes. Clones heterozygous for a dominant allele in the macronucleus produce sublines stably manifesting only one of the alleles (phenotypic assortment). Phenotypic assortment seems to be a random process, with a low rate of fixation for a particular allele. The observed coassortment of secretion deficiency and the development of the vacuolarized phenotype support our assumption that both traits are caused by the same mutational block and that this block is due to a recessive single gene mutation a t the sec locus. Analysis of F, and test cross progeny. These assumptions were further analyzed by crossing the two F, clones 9a and l l a to MS-1 and to each other (Table 3). The results supported the model that the block in secretion of acid hydrolases in the mutant MS-1 is caused by a single gene mutation at a locus we named sec. The recessive mutation was designated sec. As expected, all sec progeny clones (analyzed in Table 3) also developed the vacuolarized phenotype when starved in low-osmolarity buffers. Therefore, we conclude that the mutation a t the sec locus is also responsible for the vacuolarized phenotype.
Chromosomal Location of the sec locus of MS-1 by Deletion Mapping The technique of deletion mapping using nullisomic heterokaryons has been developed and described by Bruns and Brussard [1981] and Bruns et al. [1983]. The nullisomic strains have a normal macronucleus and therefore are viable, although one or more entire chromosome pair is missing in the micronucleus. We crossed the mutant MS-1 to different nullisomic strains (all wild type with respect to secretion of acid hydrolases; Table 4).If the see locus is not on the missing chromosome, progeny clones will secrete acid hydrolases with wild-type characteristics (see Table 4);if it is however on the missing chromosome, the progeny will be hemizygous for the mutant allele and will express the mutant phenotype. From the results shown in Table 5, it can be concluded that the sec locus is located on chromosome 4. All 27 progeny clones of the cross CU 383 x MS-1 were blocked in secretion of acid hydrolases and developed the vacuolarized phenotype.
DISCUSSION In this study, we established the genetic basis of secretion deficiency in the sec mutant MS-1 of Tetrahymenu thermophila. We have shown that the block in secretion and expression of the vacuolarized phenotype are most probably controlled by the same sec allele.
GENETICS OF THE SECRETORY MUTANT MS-1
171
Fig. 2. Thin section through the mutant MS-1 starved for 1 hr in Dryl's solution. Autophagous vacuoles (e.g., V) a t different stages of digestion are present. The thick arrow points to a phagocytosed mitochondrium and the thin arrows to a tiny dictyosome at the periphery of the cell. Bar = 1 Fm.
TABLE 2. Genetic Analysis of MS-1: Identification of True F, Progeny of the Cross CU 428 X MS-1
Isolated pairs 286
Viable svnclones
Immature clones
192
12
Immature clones resistant against cycloheximide and 6-methvlpurine 9=
"All nine clones are not blocked in secretion of acid hydrolases and none developed the vacuolarized phenotype when starved in low-osmolarity buffers.
During the breeding analysis of MS-1, a major obstacle has been the very low fertility of the mutant. Although we often observed a lowered fertility in mutagenized short-circuit genomic exclusion progeny, in this case i t is possible that the vacuoles emerging when the cells are transferred into Dryl's solution may have disturbed the correct transfer and movement of the pronuclei during conjugation. In spite of this, i t has been possible to select true exconjugant clones by the use of a functional heterokaryon with a different drug resistance marker. Genetic crosses of the F, clones made clear that a single gene mutation with a recessive character is responsible for the block in secretion MS-1. By these crosses and by selecting assorting clones from the het-
erozygous F,, we obtained additional evidence for our assumption t h a t the single gene mutation at the sec locus is also responsible for the abnormal vacuolarized phenotype of the mutant. The question whether this finding reflects a physiological linkage between osmoregulation and secretion of acid hydrolases is not answered. When size and pulsation frequency of the contractile vacuole in the mutant and wild-type cells were compared, no significant differences were found. This indicates that the function of the contractile vacuole is not affected by the mutation. The nature of the vacuolarization emerging when mutant cells are starved in low-osmolarity buffers remains to be studied. Ultrastructurally, the vacuoles at least at the beginning of starvation resemble autophagous vacuoles containing mitochondria. The number of these autophagous vacuoles seems to be higher in the mutant, so small breakdown products may build up a high osmotic pressure in these vacuoles. In the wild-type, due to the secretion of large amounts of acid hydrolases, a more balanced situation may exist. Through the use of specific antibodies against two major lysosomal enzymes, P-hexosaminidase and acid phosphatase, we are trying to localize these enzymes in the vacuoles. The nullisomic strains [Bruns and Brussard, 1981;
172
HUNSELER AND TIEDTKE TABLE 3. Genetic Analysis of MS-1: Identification of Progeny of Test Crosses and Sib Crosses
Cross 9a x MS-1
Viable synclonesl isolated pairs 1881528
l l a x MS-1 Theoretical ratio 9a x l l a
Viable synclonesl sexually immature clones 59 (21)" . .
401440
7 (7)"
Phenotypes of progeny CYcloheximide G-MethYlPUrine Res. Sens. Res. Sens. 0 15 6 21
P 7
0 Al1:Res.
27
62 (30)"
2401484
Theoretical ratio
2
0.05
Al1:Res.
0.05
3 1:l
3
P
=
4 22
P
8 0.05 3:1
2
Secretion of acid hydrolases sec+ sec10 21 P 2 0.7 4 3 1:l 20 10 P 2 0.2 3:1
"In parentheses, the number of sexually immature clones used for phenotype characterization. The P values are from the test.
TABLE 4. Enzymatic Activities of Two Acid Hydrolases in Parental, F, and F, Progeny (mu . mg-' protein)*
Strain MS-1 cu 399 CU 428 CU 353 (nulli-5) CU 361 (nulli-3) CU 383 (nulli-4) 9a 9a ass. lla l l a ass. 5a (F, sec) MS-1
p-Hexosaminidase ExtraTotal cellular activitv 3 82 64 82 93 107 69 83 76 109 71 101 120 129 6 129 108 125 11 118 4 102 3 82
Acid phosphatase ExtraTotal cellular activity 5 135 43 102 61 136 110 173 66 121 68 106 60 121 7 172 57 123 14 173 5 150 5 135
*Mean of four experiments; standard deviation was below 5% for each value.
TABLE 5. Secretion of Acid Hydrolases in Progeny of Crosses of the sec Mutant MS-I With Nullisomic Strains
Cross MS-1 MS-1 MS-1
Number of Nullisomic progeny strain clones tested x CU 353 Nulli-5 13 x CU 361 Nulli-3 20 x CU 383 Nulli-4 27
13 20 0
same conditions suggests that the locus is on chromosome 4. Recently the locus of a n exocytosis mutant unable to discharge mucocysts has been localized to chromosome 5 [Bleyman and Satir, 19901. T . thermophila possesses secretory lysosomes and mucocysts. While the contents of the lysosomes are released continuously, those of the mucocysts are stored and are released only upon stimulation by a secretagogue. Both secretory pathways, the constitutive release of acid hydrolases and the regulated secretion of mucocyst contents, occur in the same single cell. Moreover, marker proteins [Hiinseler et al., 1988; Maihle and Satir, 19861 and specific mutants blocked in secretion are available. More of these mutants and their combinations by crosses may allow a genetic dissection of both pathways in the future. The combination of all these features is unique among animal cells, and this makes Tetrahymena a n attractive model for studies on the mechanisms that control sorting of proteins for the different secretory pathways.
ACKNOWLEDGMENTS
Secretion of acid
set+
x2
sec0 0 27
asec, extracellular enzyme activities for acid phosphatase 5 3.5 mU . mg-' and for P-hexosaminidase I7 mU . mg-'; sec , extracellular enzyme activities for acid phosphatase 2 65 mU . mg-' and for p-hexosaminidase 2 60 mU . mg-'. +
Bruns et al., 19831 enabled us to localize the sec allele to chromosome 4. Normal secretion rates were observed in crosses to the nulli-5 and nulli-3 strains, indicating t h a t cross fertilization occurred and that the sec locus is not on chromosome 5 or 3. The failure to secrete acid hydrolases in the cross to nulli 4 under the
We thank Dr. D. Nanney for many helpful and constructive discussions during the course of this study, when he was a Humboldt fellow a t our institute. Support from the Deutsche Forschungsgemeinschaft (SFB 310, A 3) is also gratefully acknowledged.
REFERENCES Bleyman LK, Bruns P J (1977):Genetics of cycloheximide resistance in Tetruhymena. Genetics 87275-284. Bleyman LK, Satir BH (1990):Chromosomal localization of a n exocytosis mutant in Tetruhymenu thermophila. J Protozool 37:471472. Blum J J , Rothstein TL (1975):Lysosomes in Tetruhymenu. In Dingle JT, Dean RT, Sly W (eds): "Lysosomes in Biology and Pathology." Amsterdam: North Holland, pp 33-45. Bruns PJ, Brussard TB (1974):Positive selection for mating with functional heterokaryons in Tetrahymena pyriformis. Genetics 78: 831-841. Bruns PJ, Brussard TB (1981):Nullisomic Tetrahymenu: eliminating germinal chromosomes. Science 213549-551. Bruns PJ, Brussard TB, Merriam EV (1983):Nullisomic Tetruhymena
GENETICS OF THE SECRETORY MUTANT MS-1 11. A set of nullisomics define the germinal chromosomes. Genetics 104:257-270. Florin-Christensen J, Florin-Christensen M, Tiedtke A, Rasmussen L (1990): The role of secreted acid hydrolases in the utilization of complex nutrients by Tetrahymena. Microb Ecol 19:311-316. Hunseler P, Scheidgen-Kleyboldt G , Tiedtke A (1987): Isolation and characterization of a mutant of Tetrahymena thermophila blocked in secretion of lysosomal enzymes. J Cell Sci 88:47-55. Hunseler P, Tiedtke A, von Figura K (1988): Biosynthesis of secreted p-hexosaminidase in Tetrahymenu thermophila. A comparison of the wild type with a secretory mutant. Biochem J 252837-842. Maihle NJ, Satir BH (1986): Protein secretion in Tetruhymena thermopila. Characterization of the major proteinaceous secretory proteins. J Biol Chem 261:7566-7570. Muller M (1972): Secretion of acid hydrolases and its intracellular source in Tetruhymenu pyriformis. J Cell Biol 52:478-487. Orias E, Bruns PJ (1976): Production and isolation of mutants in Tetrahymena. In Prescott DM (ed): “Methods in Cell Biology.” New York: Academic Press, Vol 13, pp 247-282.
173
Orias E, Flacks M, Satir BH (1983): Isolation and ultrastructural characterization of secretory mutants of Tetruhymenu thermophila. J Cell Sci 64:49-67. Tiedtke A (1983): Purification and properties of secreted N-acety-pD-hexosaminidase of Tetrahymena thermophila. Comp Biochem Physiol 75B:239-249. Tiedtke A, Hunseler P, Florin-Christensen J, Florin-Christensen M (1988a): Exocytosis, endocytosis and membrane recycling in Tetrahymena thermophilu. J Cell Sci 89:515-520. Tiedtke A, Rasmussen L (1988): Lysosomal enzymes in extracellular digestion in the unicellular eukaryote Tetrahymena. J Cell Physiol 136554-556. Tiedtke A, Rasmussen L (1989):Constitutive secretion of acid hydrolases in Tetrahymenu thermophila. J Protozool 36:378-382. Tiedtke A, Rasmussen L, Florin-Christensen J, Florin-Christensen M (1988b): Release of lysosomal enzymes in Tetrahymenu; A Ca2+dependent secretory process. J Cell Sci 89:167-171.
Genetic Characterization of the Secretory Mutant MS-1 of Tetrahymena thermophila: Vacuolarization and Block in Secretion of Lysosomal Hydrolases Are Caused by a Single Gene Mutation PETER HUNSELER AND ARNO TIEDTKE Institute of General Zoology and Genetics, University of Munster, Munster, Federal Republic of Germany
ABSTRACT The genetics and phenotypic features (light and electron microscopy) of a secretory mutant, MS-1, of Tetrahyrnena therrnophila blocked in secretion of lysosomal acid hydrolases have been analyzed. Although blocked constitutively in secretion, MS-1 contains active lysosomal hydrolases in amounts equivalent to the wild type. The 3:l segregation in F-2 in sib crosses and the 1 : l segregation in test crosses indicate that the block in secretion of lysosomal hydrolases is controlled by a recessive single gene locus termed sec. The sec allele of MS-1 proved also to be responsible for the highly vacuolarized phenotype the mutant developed when it was transferred from nutrient medium into buffers of low ionic strength. Deletion mapping by crossing MS-1with nullisomic strains, all secreting lysosomal hydrolases at wildtype rates, was performed. The sec phenotype was expressed in monosomic-4 progeny only, indicating that the sec allele is located on chromosome 4 of T. therrnophila. o 1992 WiIey-Liss, Inc. Key words: Acid hydrolases, sec allele, phenotype
INTRODUCTION The ciliated protozoon Tetrahymena thermophila secretes constitutively large amounts of acid hydrolases into the culture medium [Muller, 1972; Blum and Rothstein, 1975; Tiedtke and Rasmussen, 19891. Growing and multiplying cells replace the secreted hydrolases constantly, while cells suspended in starvation medium cannot replace the secreted enzymes and may lose up to 80% of the enzyme activities initially present in the cells. We recently established a screening procedure for isolation of mutants with blocks in secretion of lysosomal enzymes in order to get tools for a genetic dissection of the pathway these secreted enzymes take through the cell. The secretory mutant MS-1 dealt with
0 1992 WILEY-LISS, INC.
in this study is constitutively blocked in secretion of lysosomal acid hydrolases. It contains intracellular amounts of hydrolases similar to those in wild-type cells [Hunseler et al., 19871. Studies on the biosynthesis of two lysosomal hydrolases, P-hexosaminidase [Hunseler et al., 19881and a-glucosidase (unpublished data), have shown no detectable differences in synthesis and processing of these enzymes in mutant and wild-type cells. The mutant grows on a diet of bacteria as well as do wild-type cells, indicating that it contains intracellularly fully functional lysosomes. We therefore assume that a late step in the secretory pathway of lysosomal acid hydrolases is affected by the mutation [Tiedtke et al., 1988al. The sec mutant MS-1 has already become a valuable tool in studies on the biological significance of extracellular acid hydrolases [Tiedtke and Rasmussen, 1988; Florin-Christensen et al., 19901 and on the control of release of acid hydrolases [Tiedtke et al., 1988bl. Studies on sec and ex0 mutants (the latter blocked in secretion of mucocysts [Orias et al., 19831revealed that Tetrahymena possesses at least two independent secretory pathways: the regulated exocytosis of mucocysts and the constitutive secretion of acid hydrolases. While a n ex0 mutant (SB 2811, possessing no mature mucocysts, released acid hydrolases to the same extent as wild-type cells, MS-1 nonetheless is able to extrude mucocysts [Tiedtke et al., 1988a, and unpublished observations]. Upon starvation in inorganic buffers of low ionic strength, MS-1 develops numerous translucent vacuoles, which progressively enlarge and finally fuse to
Received for publication October 7, 1991; accepted November 11, 1991. Address reprint requests to Dr. Arno Tiedtke, Institute for General Zoology and Genetics, Schlopplatz 5, D-4400Miinster, Federal Republic of Germany.
168
HUNSELER AND TIEDTKE -
Strain cu 399 CU 428 cu 353 CU 361
cu 383
TABLE 1. Tetrahyrnena Strains Micronuclear genotype ChxlChx MprlMpr Nulli-5; MprlMpr Nulli-3; ChxlChx; MprlMpr Nulli-4; ChxlChx; MprIMpr
form a huge vacuole that dramatically changes the morphology of the cells. In the present study, we characterized genetically the expression and transmission of the sec allele and show that the block in secretion of lysosomal hydrolases and the change in cell shape are caused by the same single gene mutation.
Macronuclear phenotype Cycloheximide sensitive 6-Methylpurine sensitive 6-Methylpurine sensitive Cycloheximide sensitive; 6-Methylpurine sensitive Cycloheximide sensitive; 6-Methylpurine sensitive
Mating type
IV VII IV IV
VI
Drugs Drug resistance of progeny was tested in PPYS/PS medium containing 6-methylpurine and/or cycloheximide in final concentrations of 15 pg X m1-I. Crosses Standard crosses were carried out under axenic conditions in Dryl’s solution principally as described by Orias and Bruns [1976]. Pairs were isolated with micropipettes into drops of PPYS/PS medium on Petri dishes and allowed to grow at 30°C for 48 hr. Synclones were directly replicated into microtiter plates, containing 0.1 ml medium per well.
MATERIALS AND METHODS Strains The secretory mutant MS-1 of T. thermophila, blocked constitutively in secretion of lysosomal hydrolases, was obtained from progeny of clone CU 399, previously subjected to mutagenesis with 10 pg x ml-’ N-methyl-N’-nitro-N-nitrosoguanidine, mated with Identification of True Progeny clone C* followed by short-circuit genomic exclusion with positive selection for mating [Bruns and BrusTrue progeny are clones that possess a newly develsard, 19741. The genetic marker being used for the se- oped macronucleus after conjugation. In many cases, at lection for mating was a dominant allele (Chx)confer- least one of the parental clones was a functional hetring resistance to cycloheximide [Bleyman and Bruns, erokaryon, phenotypically sensitive to cycloheximide 19771. Owing to the nature of short-circuit genomic or 6-methylpurine. True progeny clones therefore could exclusion, the secretory mutant may be presumed to be be identified by their newly acquired resistance to homozygous for the Chx allele as well as for the allele these drugs. In rare cases when these markers were conferring the block in secretion of lysosomal hydro- absent, true progeny clones were identified by their lases. The screening for mutants blocked in secretion of sexual immaturity, a s indicated by a mating test (malysosomal hydrolases has been reported elsewhere turity test) as described below. [Hunseler et al., 19871. Clones that did not conjugate retain the old macroThe following stocks were used in the genetic anal- nucleus and remain sexually mature. Clones were ysis of MS-1 (Table 1).Following the nomenclature pro- tested against at least two nonparental mating types posed by Bruns and Brussard [1974], CU 383, for ex- and were considered immature and were retained for ample, has a micronuclear genotype homozygous further work if they did not form pairs. Immature resistant to cycloheximide and 6-methylpurine; in ad- clones were transferred until maturity was reached dition, chromosome 4 is missing in the micronucleus. and accepted as true progeny by the most stringent In the macronuclear phenotype, the strain is sensitive criterion of having nonparental mating types. to both drugs and expresses mating type VI. A collection of mating type tester strains (mating types I-VII) Mass Selection of Progeny and the functional heterokaryons were donated by Dr. In crosses with nullisomic strains, small samples (10 P. Bruns (Cornell University, Ithaca, NY). p1) of the mating mixture were distributed 12 h r after mixing of the cells to wells of microtiter plates containMedia ing 90 p1 of PPYS/PS medium per well. Cycloheximide Stock cultures were maintained in test tubes on 1% and 6-methylpurine dissolved in the medium at 30 pg proteose peptone (Difco). Mass cultures were grown a t ml-’ each were added in equivalent volumes to the 30°C on 1% proteose peptone supplemented with 0.1% wells. Nonconjugant clones, being sensitive to the yeast extract (Difco) 0.003% sequestrene (Geigy) and drugs, are killed by 3 days in the presence of the drugs. 250 pg x ml-’ each of penicillin G and streptomycin Surviving cells were then removed from the wells and cloned. The clones were grown in the presence of both sulphate (Sigma) (abbreviated PPYS/PS medium).
GENETICS OF THE SECRETORY MUTANT MS-1
169
drugs in final concentrations of 15 pg x ml-' each and were tested for vacuolarization and secretion of lysosoma1 hydrolases.
uranyl acetate and examined with a Siemens electron microscope.
Mating Tests Mating tests were performed immediately after conjugation to identify potential true exconjugant clones. Exconjugant clones and strains of mating type testers were coreplicated into V-bottom microtiter plates, grown overnight a t 30°C, and washed into Dryl's solution so that pair formation could take place in the same well. Failure of pair formation with any of the tester strains in four independent experiments was taken as indication of immature true exconjugant clones. Mating tests were repeated after exconjugant clones had matured to determine the mating type of the clone by testing it against all mating type testers.
RESULTS Morphological Characteristics of the sec Mutant MS 1 MS-1 is blocked in secretion of lysosomal acid hydrolases and releases less than 5% of its total enzyme activities into the extracellular fluids [Hunseler et al., 19871. When starved in Dryl's solution or other lowmolarity buffers for 4-6 h r to initiate conjugation, the mutant exhibits a dramatic change in its morphology (Figs. l a and 2): translucent vacuoles fuse and enlarge progressively until the cells contain one or two huge vacuoles. Wild-type cells starved for the same time become slim and contain only a few small vacuoles (Fig. lb). Unlike the mutant, wild-type cells had released by the same time 80% of their initial amount of @-hexosaminidase and 40-50% of acid phosphatase activities to the starvation medium. With the breeding analysis of the see allele we therefore also scored for the vacuolarized phenotype of progeny clones to test whether the sec- and the vacuolarized phenotypes are coexpressed or independently transferred to progeny clones.
Test for Secretion of Lysosomal Hydrolases and Vacuolarized Progeny Five milliliter samples of progeny clones were grown to cell concentrations of 5 x lo5 cells x ml-', washed twice with and resuspended in 5 ml of Dryl's solution, and kept for 4 h r a t 30°C. The appearance of vacuolarized progeny clones was scored. The extracellular and total activities of lysosomal hydrolases were determined a s follows; one-half of a sample was layered on a cushion of 10% Ficoll (Pharmacia) and centrifuged for 3 min a t 600 g. The supernatant was removed and assayed for the extracellular (secreted) lysosomal enzyme activities. The other one-half of the sample was frozen and thawed and then sonicated on ice for 30 sec a t 150 mA (Bandelin Sonorex). Particulate materials were sedimented at 10,000 g for 4 min (Runne model 85-1 microcentrifuge). The clarified supernatant was used to measure the total activities of lysosomal enzymes, i.e., cellular and extracellular activities, and to determine the protein concentration by the method of Lowry et al. [19511. The activities of N-acetyl-P-D-hexosaminidase (EC 3.2.1.52 = 6-hexosaminidase) and acid phosphatase (EC 3.1.3.2) were determined as described by Tiedtke [1983]. One unit was defined a s the hydrolase activity that releases 1 pmol p-nitrophenollmin at 37°C. Progeny clones were regarded a s blocked in secretion of lysosomal hydrolases if the secreted activities of both enzymes were less than 10% of those found in the total system. Light and Electron Microscopy (EM) Cells were washed twice with 50 mM PIPES buffer, pH 7.2; resuspended in a small volume of the buffer; and fixed with glutaraldehyde at a final concentration of 2% for 30 min at room temperature. Light micrographs were taken with phase optics using a Zeiss photomicroscope. For EM, the cells were postfixed with 1.5% OsO, for 30 min a t room temperature, washed, dehydrated in a n ethanol series, and embedded in Epon. Thin sections were stained with lead citrate and
Genetic Analysis of the Mutant MS-1 Blocked in Secretion of Acid Hydrolases Analysis of F, progeny. Due to the nature of shortcircuit genomic exclusion, the cycloheximide-resistant mutant MS-1 was presumed to be homozygous for the Chx allele and, as a whole-genome homozygous ( = isozygous) clone, necessarily also homozygous for the allele(s), conferring the block in secretion of acid hydrolases and the expression of the vacuolarized phenotype. To test this assumption, we crossed MS-1 (previously shown to express mating type 11) to CU 428. From this cross, conjugant pairs were isolated and the synclones tested for sexual immaturity. From sexually immature synclones, single cells were cloned and tested for resistance against both drugs, cycloheximide and 6-methylpurine. Only nine of 192 viable synclones fulfilled these criteria and were regarded as true F, exconjugants. The nine clones secreted acid hydrolases in amounts similar to their parent strain, CU 399, and none of them developed the vacuolarized phenotype of MS-1. The allele(s) conferring the block in secretion of acid hydrolases and expression of the vacuolarized phenotype behaved recessively (Table 2). The genetic constitution of MS-1 therefore was assumed to be seclsec, ChxlChx (cycloheximide resistant 11). Phenotypic assortment of F, progeny. The nine heterozygous F, clones were subcloned daily and tested for sexual maturity. Two clones with mating types different from both parent strains, 9a and l l a , were chosen for further analysis. In the course of maturity tests,
170
HUNSELER AND TIEDTKE
Fig. 1. a,b Phase-contrast micrographs of cells starved for 3 h r in Dryl’s solution. Highly translucent vacuoles are observed only in the sec mutant MS-1(a); the wild type contains only a few small vacuoles (b). Bar = 10 p m
a few cells were observed in these clones showing the vacuolarized phenotype of MS-1. These cells were subcloned until they uniformly expressed the vacuolarized phenotype. When these assorters, 9a ass. and l l a ass., were tested for secretion of acid hydrolases, they were found to be unable to secrete the enzymes. Clones heterozygous for a dominant allele in the macronucleus produce sublines stably manifesting only one of the alleles (phenotypic assortment). Phenotypic assortment seems to be a random process, with a low rate of fixation for a particular allele. The observed coassortment of secretion deficiency and the development of the vacuolarized phenotype support our assumption that both traits are caused by the same mutational block and that this block is due to a recessive single gene mutation a t the sec locus. Analysis of F, and test cross progeny. These assumptions were further analyzed by crossing the two F, clones 9a and l l a to MS-1 and to each other (Table 3). The results supported the model that the block in secretion of acid hydrolases in the mutant MS-1 is caused by a single gene mutation at a locus we named sec. The recessive mutation was designated sec. As expected, all sec progeny clones (analyzed in Table 3) also developed the vacuolarized phenotype when starved in low-osmolarity buffers. Therefore, we conclude that the mutation a t the sec locus is also responsible for the vacuolarized phenotype.
Chromosomal Location of the sec locus of MS-1 by Deletion Mapping The technique of deletion mapping using nullisomic heterokaryons has been developed and described by Bruns and Brussard [1981] and Bruns et al. [1983]. The nullisomic strains have a normal macronucleus and therefore are viable, although one or more entire chromosome pair is missing in the micronucleus. We crossed the mutant MS-1 to different nullisomic strains (all wild type with respect to secretion of acid hydrolases; Table 4).If the see locus is not on the missing chromosome, progeny clones will secrete acid hydrolases with wild-type characteristics (see Table 4);if it is however on the missing chromosome, the progeny will be hemizygous for the mutant allele and will express the mutant phenotype. From the results shown in Table 5, it can be concluded that the sec locus is located on chromosome 4. All 27 progeny clones of the cross CU 383 x MS-1 were blocked in secretion of acid hydrolases and developed the vacuolarized phenotype.
DISCUSSION In this study, we established the genetic basis of secretion deficiency in the sec mutant MS-1 of Tetrahymenu thermophila. We have shown that the block in secretion and expression of the vacuolarized phenotype are most probably controlled by the same sec allele.
GENETICS OF THE SECRETORY MUTANT MS-1
171
Fig. 2. Thin section through the mutant MS-1 starved for 1 hr in Dryl's solution. Autophagous vacuoles (e.g., V) a t different stages of digestion are present. The thick arrow points to a phagocytosed mitochondrium and the thin arrows to a tiny dictyosome at the periphery of the cell. Bar = 1 Fm.
TABLE 2. Genetic Analysis of MS-1: Identification of True F, Progeny of the Cross CU 428 X MS-1
Isolated pairs 286
Viable svnclones
Immature clones
192
12
Immature clones resistant against cycloheximide and 6-methvlpurine 9=
"All nine clones are not blocked in secretion of acid hydrolases and none developed the vacuolarized phenotype when starved in low-osmolarity buffers.
During the breeding analysis of MS-1, a major obstacle has been the very low fertility of the mutant. Although we often observed a lowered fertility in mutagenized short-circuit genomic exclusion progeny, in this case i t is possible that the vacuoles emerging when the cells are transferred into Dryl's solution may have disturbed the correct transfer and movement of the pronuclei during conjugation. In spite of this, i t has been possible to select true exconjugant clones by the use of a functional heterokaryon with a different drug resistance marker. Genetic crosses of the F, clones made clear that a single gene mutation with a recessive character is responsible for the block in secretion MS-1. By these crosses and by selecting assorting clones from the het-
erozygous F,, we obtained additional evidence for our assumption t h a t the single gene mutation at the sec locus is also responsible for the abnormal vacuolarized phenotype of the mutant. The question whether this finding reflects a physiological linkage between osmoregulation and secretion of acid hydrolases is not answered. When size and pulsation frequency of the contractile vacuole in the mutant and wild-type cells were compared, no significant differences were found. This indicates that the function of the contractile vacuole is not affected by the mutation. The nature of the vacuolarization emerging when mutant cells are starved in low-osmolarity buffers remains to be studied. Ultrastructurally, the vacuoles at least at the beginning of starvation resemble autophagous vacuoles containing mitochondria. The number of these autophagous vacuoles seems to be higher in the mutant, so small breakdown products may build up a high osmotic pressure in these vacuoles. In the wild-type, due to the secretion of large amounts of acid hydrolases, a more balanced situation may exist. Through the use of specific antibodies against two major lysosomal enzymes, P-hexosaminidase and acid phosphatase, we are trying to localize these enzymes in the vacuoles. The nullisomic strains [Bruns and Brussard, 1981;
172
HUNSELER AND TIEDTKE TABLE 3. Genetic Analysis of MS-1: Identification of Progeny of Test Crosses and Sib Crosses
Cross 9a x MS-1
Viable synclonesl isolated pairs 1881528
l l a x MS-1 Theoretical ratio 9a x l l a
Viable synclonesl sexually immature clones 59 (21)" . .
401440
7 (7)"
Phenotypes of progeny CYcloheximide G-MethYlPUrine Res. Sens. Res. Sens. 0 15 6 21
P 7
0 Al1:Res.
27
62 (30)"
2401484
Theoretical ratio
2
0.05
Al1:Res.
0.05
3 1:l
3
P
=
4 22
P
8 0.05 3:1
2
Secretion of acid hydrolases sec+ sec10 21 P 2 0.7 4 3 1:l 20 10 P 2 0.2 3:1
"In parentheses, the number of sexually immature clones used for phenotype characterization. The P values are from the test.
TABLE 4. Enzymatic Activities of Two Acid Hydrolases in Parental, F, and F, Progeny (mu . mg-' protein)*
Strain MS-1 cu 399 CU 428 CU 353 (nulli-5) CU 361 (nulli-3) CU 383 (nulli-4) 9a 9a ass. lla l l a ass. 5a (F, sec) MS-1
p-Hexosaminidase ExtraTotal cellular activitv 3 82 64 82 93 107 69 83 76 109 71 101 120 129 6 129 108 125 11 118 4 102 3 82
Acid phosphatase ExtraTotal cellular activity 5 135 43 102 61 136 110 173 66 121 68 106 60 121 7 172 57 123 14 173 5 150 5 135
*Mean of four experiments; standard deviation was below 5% for each value.
TABLE 5. Secretion of Acid Hydrolases in Progeny of Crosses of the sec Mutant MS-I With Nullisomic Strains
Cross MS-1 MS-1 MS-1
Number of Nullisomic progeny strain clones tested x CU 353 Nulli-5 13 x CU 361 Nulli-3 20 x CU 383 Nulli-4 27
13 20 0
same conditions suggests that the locus is on chromosome 4. Recently the locus of a n exocytosis mutant unable to discharge mucocysts has been localized to chromosome 5 [Bleyman and Satir, 19901. T . thermophila possesses secretory lysosomes and mucocysts. While the contents of the lysosomes are released continuously, those of the mucocysts are stored and are released only upon stimulation by a secretagogue. Both secretory pathways, the constitutive release of acid hydrolases and the regulated secretion of mucocyst contents, occur in the same single cell. Moreover, marker proteins [Hiinseler et al., 1988; Maihle and Satir, 19861 and specific mutants blocked in secretion are available. More of these mutants and their combinations by crosses may allow a genetic dissection of both pathways in the future. The combination of all these features is unique among animal cells, and this makes Tetrahymena a n attractive model for studies on the mechanisms that control sorting of proteins for the different secretory pathways.
ACKNOWLEDGMENTS
Secretion of acid
set+
x2
sec0 0 27
asec, extracellular enzyme activities for acid phosphatase 5 3.5 mU . mg-' and for P-hexosaminidase I7 mU . mg-'; sec , extracellular enzyme activities for acid phosphatase 2 65 mU . mg-' and for p-hexosaminidase 2 60 mU . mg-'. +
Bruns et al., 19831 enabled us to localize the sec allele to chromosome 4. Normal secretion rates were observed in crosses to the nulli-5 and nulli-3 strains, indicating t h a t cross fertilization occurred and that the sec locus is not on chromosome 5 or 3. The failure to secrete acid hydrolases in the cross to nulli 4 under the
We thank Dr. D. Nanney for many helpful and constructive discussions during the course of this study, when he was a Humboldt fellow a t our institute. Support from the Deutsche Forschungsgemeinschaft (SFB 310, A 3) is also gratefully acknowledged.
REFERENCES Bleyman LK, Bruns P J (1977):Genetics of cycloheximide resistance in Tetruhymena. Genetics 87275-284. Bleyman LK, Satir BH (1990):Chromosomal localization of a n exocytosis mutant in Tetruhymenu thermophila. J Protozool 37:471472. Blum J J , Rothstein TL (1975):Lysosomes in Tetruhymenu. In Dingle JT, Dean RT, Sly W (eds): "Lysosomes in Biology and Pathology." Amsterdam: North Holland, pp 33-45. Bruns PJ, Brussard TB (1974):Positive selection for mating with functional heterokaryons in Tetrahymena pyriformis. Genetics 78: 831-841. Bruns PJ, Brussard TB (1981):Nullisomic Tetrahymenu: eliminating germinal chromosomes. Science 213549-551. Bruns PJ, Brussard TB, Merriam EV (1983):Nullisomic Tetruhymena
GENETICS OF THE SECRETORY MUTANT MS-1 11. A set of nullisomics define the germinal chromosomes. Genetics 104:257-270. Florin-Christensen J, Florin-Christensen M, Tiedtke A, Rasmussen L (1990): The role of secreted acid hydrolases in the utilization of complex nutrients by Tetrahymena. Microb Ecol 19:311-316. Hunseler P, Scheidgen-Kleyboldt G , Tiedtke A (1987): Isolation and characterization of a mutant of Tetrahymena thermophila blocked in secretion of lysosomal enzymes. J Cell Sci 88:47-55. Hunseler P, Tiedtke A, von Figura K (1988): Biosynthesis of secreted p-hexosaminidase in Tetrahymenu thermophila. A comparison of the wild type with a secretory mutant. Biochem J 252837-842. Maihle NJ, Satir BH (1986): Protein secretion in Tetruhymena thermopila. Characterization of the major proteinaceous secretory proteins. J Biol Chem 261:7566-7570. Muller M (1972): Secretion of acid hydrolases and its intracellular source in Tetruhymenu pyriformis. J Cell Biol 52:478-487. Orias E, Bruns PJ (1976): Production and isolation of mutants in Tetrahymena. In Prescott DM (ed): “Methods in Cell Biology.” New York: Academic Press, Vol 13, pp 247-282.
173
Orias E, Flacks M, Satir BH (1983): Isolation and ultrastructural characterization of secretory mutants of Tetruhymenu thermophila. J Cell Sci 64:49-67. Tiedtke A (1983): Purification and properties of secreted N-acety-pD-hexosaminidase of Tetrahymena thermophila. Comp Biochem Physiol 75B:239-249. Tiedtke A, Hunseler P, Florin-Christensen J, Florin-Christensen M (1988a): Exocytosis, endocytosis and membrane recycling in Tetrahymena thermophilu. J Cell Sci 89:515-520. Tiedtke A, Rasmussen L (1988): Lysosomal enzymes in extracellular digestion in the unicellular eukaryote Tetrahymena. J Cell Physiol 136554-556. Tiedtke A, Rasmussen L (1989):Constitutive secretion of acid hydrolases in Tetrahymenu thermophila. J Protozool 36:378-382. Tiedtke A, Rasmussen L, Florin-Christensen J, Florin-Christensen M (1988b): Release of lysosomal enzymes in Tetrahymenu; A Ca2+dependent secretory process. J Cell Sci 89:167-171.
