- Email: [email protected]
A Mutational Analysis of Conjugation inTetrahymena thermophila
DEVELOPMENTAL BIOLOGY 189, 233–245 (1997) ARTICLE NO. DB978649
A Mutational Analysis of Conjugation in Tetrahymena thermophila 2. Phenotypes Affecting Middle and Late Development: Third Prezygotic Nuclear Division, Pronuclear Exchange, Pronuclear Fusion, and Postzygotic Development
Eric S. Cole*,1 and Timothy A. Soelter* *Biology Department, St. Olaf College, Northfield, Minnesota 55057
Conjugation following pair formation in Tetrahymena can be divided into three distinct sequences of events: prezygotic development, postzygotic development, and exconjugant development. The decision to proceed with postzygotic development is governed by a developmental checkpoint occurring sometime during the middle stages of conjugation. A second developmental decision is made to initiate pair separation and exconjugant development. This paper examines the phenotypes of five newly isolated conjugation mutants (cnj6–cnj10) which affect middle and late events within the conjugation program. cnj6 mutants exhibit normal nuclear behavior throughout development up to and including differentiation of new macronuclear anlagen. Pairs arrest at this developmental endpoint, unable to dissociate. cnj7 and cnj8 eliminate the third prezygotic nuclear division and the first postzygotic nuclear division. All subsequent developmental events appear normal. cnj9 eliminates the second postzygotic nuclear division, and subsequently, new macronuclei fail to develop despite parental macronuclear degradation. cnj10 results in a pleiotropic phenotype characterized by failure of numerous events which all appear to involve nuclear–cytoskeletal interactions. These defects include nuclear selection (anchoring nuclei to the exchange junction), pronuclear exchange, pronuclear fusion, and anchoring postzygotic nuclear division products to the posterior cell cortex. These mutant phenotypes are used to draw inferences regarding developmental dependencies that govern a cell’s entry into the postzygotic and exconjugant developmental programs. q 1997 Academic Press
INTRODUCTION One of the more intriguing aspects of conjugation in Tetrahymena is the functional separation of prezygotic, postzygotic, and ‘‘exconjugant’’ developmental subprograms. Prezygotic development is initiated by pair formation and includes two meiotic nuclear divisions, selection of one meiotic product to undergo a third nuclear mitosis, differentiation of a migratory and stationary pronucleus within each mating partner, exchange of migratory pronuclei, and fusion of the migratory pronucleus with the resident, stationary pronucleus, resulting in formation of a zygotic nucleus. The earliest events within this program, meiosis and nuclear ‘‘selection,’’ are described in the accompanying paper (Cole 1
To whom correspondence should be addressed. Fax: (507) 6463968. E-mail: [email protected].
et al., 1997). Although nuclear fusion formally marks the beginning of postzygotic development, it is unclear what event or conditions actually trigger this transition. What is clear is that when the appropriate conditions are not met, cells abort development, retain their parental macronuclei, and separate prematurely in a process known as genomic exclusion (Allen, 1967a,b; Allen et al., 1967; Doerder and Shabatura, 1980). Postzygotic events involve two consecutive nuclear divisions. The first postzygotic division spindle delivers nuclei into the anterior and posterior cytoplasm. The anterior nucleus migrates more posteriorly prior to the second division (Ray, 1956). The spindle of the second division delivers two of four nuclear products to the anterior cytoplasm where a program of chromosomal modifications is initiated transforming these two nuclei into macronuclear anlagen (MA; see Nanney, 1953). The two remaining nuclei are positioned in the posterior cytoplasm
0012-1606/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.
AID
DB 8649
/
6x2c$$$161
233
08-21-97 06:05:02
dbas
234
Cole and Soelter
where they appear to become ‘‘anchored’’ to the posterior cell cortex (Kaczanowski et al., 1991). These posterior nuclei differentiate into germ-line micronuclei. Macronuclear anlagen formation is the second case in which the anterior (paroral) cytoplasm appears to exert a determinative influence over nuclear fate. During prezygotic development, the anteriormost, paroral nucleus is ‘‘selected’’ to undergo a third, prezygotic division (see accompanying manuscript, Cole et al., 1997). In Paramecium, the situation is different in that the oral cavity is situated in the posterior third of the cell. Perhaps for this reason, it is the posterior nuclear products which differentiate into macronuclear anlagen within this species (Sonneborn, 1954). In both species, differentiation of somatic and germinal nuclei is associated with extreme polar localization, a phenomenon reminiscent of pole cell nuclear determination in the embryogenesis of Drosophila (as pointed out by Grandchamp and Beisson, 1981). It has been suggested that any nuclei migrating into the anterior cytoplasm (in Tetrahymena) will differentiate as macronuclei, but anchoring of micronuclei at the cell’s posterior cortex is necessary to protect them from macronuclear differentiation and preserve their germinal, micronuclear character (Kaczanowski et al., 1991). Macronuclear differentiation involves chromosome fragmentation, DNA elimination, telomere synthesis, and gene amplification (Yao, 1990; Blackburn, 1991; Gray et al., 1991). Eight to ten hours after cell mixing, transcription is detected in the newly formed macronuclear anlagen (Wenkert and Allis, 1984). Twelve hours after mixing (at 307C), if all nuclear events occur normally, pairs dissociate, the parental macronucleus is destroyed, and one of the two micronuclei is eliminated. Micronuclear elimination and macronuclear resorption are accompanied by regeneration of the oral apparatus (oral replacement) at approximately 14 hr (Cole and Frankel, 1991; Kiersnowska and Kaczanowski, 1993). These events will be referred to here as ‘‘exconjugant’’ development. This investigation examined five novel conjugation mutants, cnj6–cnj10, which, considered with two previously characterized mutants, bcd and janA (Cole, 1991; Cole and Frankel, 1991), affect each of the middle and late stages of nuclear behavior. In particular we have used the mutant phenotypes to shed light on developmental decisions which govern the transition from prezygotic to postzygotic and from postzygotic to exconjugant development.
METHODS AND MATERIALS The methods are described in the accompanying paper (Cole et al., 1997).
RESULTS Middle and Late Stages of Conjugation in Tetrahymena thermophila Figure 1 shows a schematic diagram of Tetrahymena conjugation. Figure 2 shows DAPI-stained representatives of
the developmental events beginning with the gametogenic (third) nuclear division (Figs. 2A and 2B) and ending with the three exconjugant stages (Figs. 2M, 2N, and 2O).
Phenotypic Profiles of Conjugation Mutants cnj7 and cnj8. cnj7 and cnj8 were originally isolated based upon quite similar terminal pair configurations with abnormal numbers of macronuclear anlagen and micronuclei. The most frequent endpoint consisted of a single, enlarged macronuclear anlagen and a single, enlarged micronucleus (Fig. 3B). Developmental analysis revealed that the earliest abnormality appeared after the completion of meiosis II (Fig. 4). At this point, the third prezygotic (gametogenic) division failed to occur. Normally, we would predict that the four haploid products would subsequently degenerate, the cells would become ‘‘starlike,’’ and development would abort following the genomic exclusion pathway (Allen, 1967a,b; Allen et al., 1967; Doerder and Shabatura, 1980). However, in cnj7 and cnj8, one of the four meiotic nuclei became associated with the exchange junction (Fig. 3A) and behaved as a migratory pronucleus while a second meiotic product appeared to differentiate as a stationary pronucleus. Both of these nuclei were decorated with fenestrin (Figs. 5A and 5B), a protein associated with differentiated pronuclei (Nelsen et al., 1994). Migratory pronuclei appeared to be exchanged, and synkarya were observed, indicating that pronuclear fusion (karyogamy) had occurred. Subsequent to nuclear fusion, there was a frequent failure of one of the two postzygotic mitoses. The most abundant endpoint possessed both a single micronucleus and a single MA, suggesting that the first division had failed and the second postzygotic division had occurred, thereby delivering one nucleus to the anterior cytoplasm (where it differentiated into a macronuclear anlagen) and one nucleus to the posterior cytoplasm where it became anchored and differentiated as a germinal micronucleus. Our temporal profile of (cnj7 1 cnj7) was compared to the wild-type profile shown in the accompanying paper (Cole et al., 1997) to distinguish whether the skipped nuclear division was the first or second postzygotic division. (Note. The cnj8 profile was very similar to the cnj7 profile and hence is not shown.) Exact timing of developmental events varies from experiment to experiment. Hence, a good internal marker is the developmental appearance of the pronuclear exchange configuration (Fig. 1J). From the wild-type profile, we saw that pairs which completed the first postzygotic division appeared 12 hr following the appearance of pairs in the exchange configuration, whereas second postzygotic division figures appeared 1 hr after the exchange configuration. From Fig. 5, we see that cnj7 pairs exhibiting the exchange configuration first appeared 4.5 – 5 hr after cell mixing. Pairs exhibiting two postzygotic nuclei appeared an hour or more after first appearance of ex-
Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.
AID
DB 8649
/
6x2c$$$162
08-21-97 06:05:02
dbas
235
Mutational Analysis of Tetrahymena Conjugation, 2
FIG. 1. A schematic diagram showing all of conjugal development in Tetrahymena thermophila.
change configurations. These results strongly suggest that cnj7 (and cnj8) pairs skip the first postzygotic division and resume development with the second postzygotic division. A heterotypic mating between (cnj7 1 cnj8) revealed complementation and a fairly high degree of cytoplasmic rescue of the conjugal phenotype. This strongly suggests that these mutations, despite their similar phenotypes, do in fact define distinct loci. cnj9. cnj9 matings exhibited perfectly normal chromosomal and nuclear behavior up to the second postzygotic mitosis (Fig. 6). This final nuclear division apparently failed in this phenotype with intriguing consequences. The parental macronucleus condensed on schedule, and yet, neither of the two postzygotic nuclei differentiated into macronuclear anlagen (see Figs. 3C and 3D). Consequently, the most common endpoint for cnj9 was a pair of cells, each with a condensed parental macronucleus and two (enlarged) micronuclei. It should be added that penetrance was not 100%. Approximately 5% of pairs showed macronuclear anlagen formation (Fig. 3D), and it is notable that in these rare cases, the second postzygotic division was also successful. (We
have found very rare cases in which only two zygotic nuclei were present and one had differentiated into a macronuclear anlagen.) Again, developmental timing convinced us that the first postzygotic nuclear division occurred on schedule, whereas the second postzygotic division failed. In (cnj9 1 cnj9) matings, pairs exhibiting two postzygotic nuclei appeared just 12 hr after first appearance of exchange configurations (Fig. 6). cnj10. The cnj10 mutant was initially isolated based upon a terminal phenotype in which pairs contained multiple macronuclear anlagen. Cells with as many as eight nuclear figures were observed (Fig. 3F). Cells homozygous for the cnj10 mutation paired normally and appeared to undergo normal development up to the third prezygotic division (see developmental profile, Fig. 7). At this point, there was a high incidence of pairs in which at least one partner eliminated all four meiotic products. In effect, nuclear selection failed quite frequently. Analysis of asymmetric pairs in which one partner exhibited nuclear selection and the other eliminated all four meiotic products was quite revealing. In such pairs it was possible to observe whether the
Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.
AID
DB 8649
/
6x2c$$$162
08-21-97 06:05:02
dbas
236
Cole and Soelter
08-21-97 06:05:02
dbas
237
Mutational Analysis of Tetrahymena Conjugation, 2
FIG. 3. (A and B) cnj7 pairs. (A) The exchange configuration for a cnj7 pair. Note total of four nuclei per cell, one at exchange junction. (B) Typical cnj7 endpoint. (Note single enlarged MA and single enlarged micronucleus, Mic). (C and D) cnj9 1 cnj9 pairs: abnormal endpoints. (C) Most typical cnj9 endpoint showing condensed parental macronucleus (PM) and two enlarged micronuclei. (D) Rare cnj9 endpoint with one normal partner. (E and F) cnj10 pairs. (E) A cnj10 pair showing the second postzygotic division spindles (note four spindles will generate eight nuclear products). Also note the unilateral development characteristic of nuclear selection failure in left-hand partner earlier in development. (F) Typical cnj10 endpoint with eight nuclei in one partner (many or most of them developing as macronuclear anlagen).
normally developing partner could transfer its migratory pronucleus unilaterally to the defective partner. In fact, this was never observed. Cells which completed the third prezygotic division were never observed to adopt the ‘‘exchange configuration’’ (see Fig. 2C). Instead, pronuclei maintained their roundness and their distance from the exchange junction. Subsequently, even in pairs in which both partners developed migratory and stationary pronuclei, synkaryon formation was never observed and the two pronuclei per-
sisted as distinct entities. The three ‘‘relic’’ nuclei were eliminated normally. In cell partners which successfully completed the third prezygotic nuclear division, the two persistent pronuclei underwent two postzygotic divisions, resulting in as many as eight mitotic products (Fig. 3F). Curiously, all of these nuclei tended to aggregate in the cytoplasm anterior to the macronucleus. Consequently, most of these nuclei initiated macronuclear anlagen development (Fig. 3F). One final observation regarding cnj10 was
FIG. 2. DAPI-stained fluorescence micrographs of middle and late stages in normal conjugal development. In each panel, nuclei from a symmetrical pair of mating cells are shown. (A) Third prezygotic division showing anaphase chromosomes (open arrow) and secondary ‘‘pseudospindle’’ (closed arrow). Densely staining small round bodies are ‘‘nuclear relics,’’ i.e., nuclei which have been targeted for elimination. (B) Completion of third prezygotic division showing still condensed mitotic chromosomes. (C) Pronuclei have assumed exchange configuration (M, migratory and S, stationary pronuclei). Note decondensed nature of pronuclei. (D) After pronuclear exchange, pronuclei have fused (arrow) forming synkaryon. (E and F) First postzygotic nuclear division (PZD). (G) Completion of first PZD. Note two rounded nuclear products. (H and I) Second PZD. (J) Completion of second PZD. Note that the anterior pair of nuclei have already assumed slightly larger dimensions. (K and L) Macronuclear anlagen differentiation. Closed arrows indicate differentiating MA. Open arrow indicates condensing parental macronucleus. (M) Pair separation. Note two enlarged macronuclear anlagen (diffuse), two brightly staining micronuclei, and the condensing parental macronucleus. (N) Parental macronucleus has been eliminated. (O) Second micronucleus has been eliminated.
Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.
AID
DB 8649
/
6x2c$$$162
08-21-97 06:05:02
dbas
238
Cole and Soelter
FIG. 4. Developmental profile for cnj7 1 cnj7. 100 pairs were scored for each time point. The y axis represents the percentage of the sample in given stages for each given time point.
that postzygotic developmental divisions were somewhat delayed. cnj6 and janA. Finally we have two mutations that affect very late stages in development. Both janA (see Cole and Frankel, 1991) and cnj6 result in a failure of pair separation after successful completion of a normal sequence of nuclear events (see also the mra mutation; Kaczanowski, 1992). These mutant cell lines apparently fail to produce some essential trigger for pair termination. In these pairs, one of the two micronuclei and the condensed, parental macronucleus failed to be resorbed at the appropriate time.
shown in Fig. 8. These phenotypes provide us with a unique opportunity to explore the dependencies and checkpoints which govern this elaborate developmental program. In particular, these mutants offer insights regarding the decision to initiate or trigger postzygotic development, the decision to initiate macronuclear anlagen formation, and the decision for pairs to separate and initiate exconjugant development (see Fig. 9 for a summary of middle and late stage developmental dependencies).
Developmental Dependencies
DISCUSSION Overview A summary of mutant phenotypes affecting middle and late stages of conjugal development in Tetrahymena is
Triggering postzygotic development. Until pronuclei fuse forming the synkaryon, mating partners are not committed to pursue the postzygotic developmental program which includes the first and second postzygotic nuclear di-
Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.
AID
DB 8649
/
6x2c$$$162
08-21-97 06:05:02
dbas
239
Mutational Analysis of Tetrahymena Conjugation, 2
FIG. 5. A cnj7 1 cnj7 mutant pair stained with both (A) 3A7, a monoclonal antiserum directed against the protein fenestrin visualized with an FITC-conjugated secondary antiserum, and (B) DAPI nuclear stain. Note fenestrin labeling of two (meiotic) nuclear products.
visions and macronuclear anlagen differentiation. In fact, pairs that run into difficulties within the early stages of development frequently retain their parental macronuclei and abort subsequent development (a process termed firstround genomic exclusion; Allen, 1967a,b; Doerder and Shabatura, 1980). There are four conspicuous candidates for the cytological trigger that commits a cell to postzygotic development: (1) completion of the third prezygotic division, (2) association of two sets of migratory pronuclei with the exchange junction, (3) pronuclear exchange, and (4) pronuclear fusion or karyogamy. Here we summarize data which suggest that none of these events is sufficient to trigger postzygotic development. Is the third prezygotic division necessary in order to trigger postzygotic development? The evidence suggests not. The third prezygotic division appears to be eliminated in both the cnj7 and the cnj8 mutants, yet postzygotic development proceeds relatively normally. Hence, the third prezygotic division is not necessary in order to trigger postzygotic development. Association of the migratory pronucleus with the exchange junction has been postulated as the event that triggers postzygotic development (Hamilton, 1984). This must happen in both partners as evidenced by genomic exclusion matings in which one partner develops normally, while the other partner eliminates all of its meiotic nuclei (Allen, 1967a,b; Doerder and Shabatura, 1980). The result is unilateral pronuclear development and transfer and abortive development; postzygotic development is not initiated. cnj10 pairs defy this hypothesis as do previously described bcd matings (Cole, 1991). cnj10 matings frequently become ‘‘unilateral’’ in that one partner destroys all of its meiotic nuclei, yet development proceeds in the partner which has successfully undergone nuclear
selection and third-prezygotic division. In fact, development proceeds (unilaterally) all the way through macronuclear anlagen formation. Hence, bilateral pronuclear association with the exchange junction is not necessary for triggering postzygotic development. This is also demonstrated by another type of aberrant mating. Uniparental cytogamy is a form of genomic exclusion mating in which self-fertilization is provoked by osmotic shock (Cole and Bruns, 1992). In these pairs, unilateral pronuclear association with the exchange junction occurs, yet cells initiate self-fertilization and postzygotic development in the normal partner. Finally, bcd matings demonstrate that bilateral pronuclear association with the exchange junction is insufficient to trigger postzygotic development. In these cells, migratory pronuclei are successfully exchanged between mating partners, yet they fail to fuse and pairs abort development (Cole, 1991). Hence, it would appear that bilateral association of pronuclei with the exchange junction is neither necessary nor sufficient to trigger postzygotic development. Pronuclear exchange can be blocked in a number of ways. Vinblastine and hyperosmotic shocks have both been shown to prevent pronuclear exchange (Hamilton and SuhrJessen, 1980; Orias et al., 1979; Orias and Hamilton, 1979), yet postzygotic development proceeds. Again, cnj10 cells also exhibit a failure of pronuclear transfer and yet development proceeds. Pronuclear fusion (karyogamy) can also be blocked using anti-microtubule drugs such as vinblastine (Hamilton, 1984; Hamilton et al., 1988) and nocodazole (Kaczanowski et al., 1991) and yet macronuclear anlagen formation and parental macronucleus condensation occur (both hallmarks of successful postzygotic development). Our cnj10 phenotype supports these findings in that cnj10 pairs exhibit pronuclear fusion failure and yet proceed with postzygotic de-
Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.
AID
DB 8649
/
6x2c$$$163
08-21-97 06:05:02
dbas
240
Cole and Soelter
FIG. 6. Developmental profile for cnj9 1 cnj9. 100 pairs were scored for each time point. The y axis represents the percentage of the sample in given stages for each given time point.
velopment. Hence, pronuclear fusion is not necessary in order to provoke postzygotic development. All these observations serve to illustrate that, so far, there is no simple, conspicuous cytological event that can be identified as the trigger for postzygotic development. This suggests that such an important developmental decision may be controlled by more than one mechanism and that it is developmentally buffered. This would make adaptive sense in that commitment to postzygotic development is also a commitment to destroying one’s functional somatic macronucleus. Further insight into the control of this developmental decision has been provided by Ward and Herrick (1996). They have shown that, whatever the cytological trigger may be, it must activate novel gene transcription and translation in order to provoke entry into postzygotic development. Triggering macronuclear anlagen formation. It has
already been well established that in Tetrahymena, micronuclei must be translocated to the anterior cytoplasm in order for them to begin differentiating into macronuclear anlagen (Nanney, 1953). Our cnj9 mutant is of interest in this regard. It is the second postzygotic division that delivers two micronuclei to the anterior cytoplasm and two to the posterior cytoplasm, and it is this same division which appears to be blocked in cnj9 matings. This is consistent with our observations that cnj9 cells also fail to differentiate macronuclear anlagen. Orias reports that vinblastine treatment during this stage results (essentially) in a phenocopy of our cnj9 mutation. He writes: ‘‘When (and only when) the last (i.e., second) postzygotic nuclear division is inhibited, macronuclear differentiation is delayed, and nuclear differentiation products intermediate in appearance between micro- and macro-nucleus are often observed (staining almost as
Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.
AID
DB 8649
/
6x2c$$$163
08-21-97 06:05:02
dbas
241
Mutational Analysis of Tetrahymena Conjugation, 2
FIG. 7. Developmental profile for cnj10 1 cnj10. 100 pairs were scored for each time point. The y axis represents the percentage of the sample in given stages for each given time point.
densely as normal new micronuclei, but larger than them in size and apparent amount of DNA)’’ (Orias, 1986, p. 69). This is a perfect description of the cnj9 phenotype. One other piece of information that our cnj9 mutant provides is that condensation of the parental macronucleus is not dependent upon macronuclear anlagen formation. Macronuclear condensation must therefore be triggered by gene products produced by this very same parental macronucleus. Similar observations have been reported by others (Nanney, 1953; Doerder and Shabatura, 1980). Macronuclear condensation is a classic example of developmentally programmed nuclear elimination. The cnj10 phenotype is also informative. The cnj10 mutation appears remarkably pleiotropic. The earliest cnj10 defect appears to be failure of nuclear selection (a prezygotic activity). A second defect appears to be an inability of the cell to transfer its migratory pronucleus
across the exchange junction. Subsequent defects include pronuclear fusion failure, postzygotic developmental delay (at least in asymmetric pairs with one starlike partner, which resemble the developmental delay reported by Gaertig and Kaczanowski, 1987), and a consequent production of supernumerary macronuclear anlagen. The supernumerary MA can be understood as the consequence of two events. First, without nuclear fusion, there are two nuclei which undergo both postzygotic divisions. This can result in as many as eight postzygotic nuclear division products. The second defect appears to be a failure of cnj10 cells to anchor the posterior nuclear division products at the posterior cortex. Kaczanowski has argued that without this cortical anchoring, nuclei migrate anteriorly and differentiate as MA (Kaczanowski et al., 1991). It is notable that at least three events involving intimate nuclear – cortical association appear defective in cnj10: se-
Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.
AID
DB 8649
/
6x2c$$$163
08-21-97 06:05:02
dbas
242
Cole and Soelter
FIG. 8. Summary of nuclear behavior during middle and late stages of conjugation and phenotypes of some conjugation mutants. Lettering refers to the stages depicted in Fig. 1.
lection of one of the four meiotic products within the protective paroral cortex, pronuclear migration across the intercellular exchange junction, and anchoring of the posterior products of the second postzygotic division to the posteriormost cortex to maintain their germinal, micronuclear character. These defects argue that cnj10 plays a role in nuclear – cytoskeletal association, thereby supporting the hypothesis that cortical anchoring protects a nucleus from various cytoplasmic signals of differentiation. Triggering pair separation and exconjugant development. Reagents that block transcription or translation, when delivered after macronuclear anlagen formation has begun, result in a pair-separation-failure syndrome (Ward and Herrick, 1996), which phenocopies the cnj6 mutant as well as janA and mrA (Cole and Frankel, 1991; Kaczanowski, 1992). This condition includes an inability of pairs to separate, failure of pairs to eliminate one of their two micronuclei, and an inability of pairs to eliminate their old condensed parental macronucleus. This condition is lethal. These results suggest that there is a sensitive developmental stage during which pairs become competent to com-
plete late conjugal and postconjugal development. It also suggests that, whereas most of development is under the genetic control of the parental macronucleus, pair separation and exconjugant development are under the genetic control of the newly formed macronuclear anlagen. We propose that these mutations (for various reasons) are unable to effect the transition from parental to zygotic nuclear gene expression. This hypothesis is currently under investigation.
Diversity of Nuclear Divisions From our studies of all 10 cnj mutants (see accompanying paper, Cole et al., 1997), the case can be made that Tetrahymena carry out seven distinct types of nuclear division (see Fig. 10). During vegetative cell growth, the micronucleus and macronucleus exhibit profoundly different types of nuclear division. The macronucleus undergoes an amitotic fission which can be functionally dissociated from the cycle of DNA synthesis (Cleffmann, 1980; Doerder and DeBault, 1978; Doerder, 1979). The micronucleus undergoes a more conventional mitosis during vegetative growth, albeit still
Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.
AID
DB 8649
/
6x2c$$$163
08-21-97 06:05:02
dbas
243
Mutational Analysis of Tetrahymena Conjugation, 2
FIG. 9. Developmental dependencies during middle and late, postzygotic conjugation in Tetrahymena. Bold arrows indicate inferred developmental dependencies. For example, pronuclear differentiation appears to be required for pronuclear exchange and pronuclear fusion. However, pronuclear fusion does not depend upon exchange (hence no arrow). Mutant designations in italics refer to those genes whose wild-type activities appear necessary for a given step. Trigger to postzygotic development, indicated by the vertical box, depends upon some as yet unidentified cytological event(s), and appears necessary to trigger an ensemble of events including the two postzygotic nuclear divisions, macronuclear anlagen differentiation, and parental macronucleus condensation. Each of these four events appears to be able to take place independently from the other three, hence they each have individual arrows driving them. It is unknown whether transcriptional activation of the macronuclear anlagen depends upon the chromosomal rearrangements that characterize macronuclear differentiation (hence the question mark). Transcriptional activation does appear to be essential for all the events associated with ‘‘exconjugant development’’ (listed under Pair Separation). Parental macronucleus condensation appears to occur independently of pair separation (hence no arrow connecting them).
within a closed nuclear envelope. During the early stages of conjugation the micronucleus undergoes three distinct types of karyokinesis: meiosis I (which involves synapsis, recombination, and absence of centromere division), meiosis II (which does exhibit centromere division), and a third ‘‘gametogenic’’ mitosis directly following a round of DNA synthesis (Doerder and Shabatura, 1980). That meiosis I and II are distinct comes as no surprise, but it is intriguing to learn that all three prezygotic divisions are distinct from one another and from the vegetative mitoses. The case for this can be made as follows. During conjugation, wild-type cnj1 and cnj2 genes are clearly essential for all three prezygotic nuclear divisions (see accompanying paper by Cole et al., 1997). Specifically, cells homozygous for the cnj1 or cnj2 mutations appear unable execute chromosome condensation; chromosomes fail to segregate, and consequently nuclei fail to divide. It would appear that cnj1 and cnj2 gene products are essential for all three prezygotic divisions. Nevertheless, the vegetative micronuclear mitosis occurs freely in cnj1 and cnj2 homozy-
gotes exhibiting normal chromatin condensation. Hence, in this way at least, vegetative micronuclear mitoses are distinct from meiosis I, meiosis II, and the third prezygotic mitosis. A complementary situation has been observed in Paramecium tetraurelia. Adl and Berger (1994) have shown that the temperature-sensitive cc1 mutation blocks micronuclear elongation (and oral development) during the vegetative cell cycle, but exhibits normal micronuclear divisions (and oral replacement) during conjugal development. The fact that cnj5 pairs skip both meiotic divisions and yet proceed through the third (gametogenic) mitosis distinguishes this third mitosis from the second meiotic division, even though both superficially resemble one another (see accompanying paper, Cole et al., 1997). The cnj7 and cnj8 mutations both appear to cause elimination of the first postzygotic mitosis but not the second. The cnj9 mutation clearly eliminates the second postzygotic mitosis but not the first. These results argue that each of these seven different nuclear divisions requires the activity of different gene products.
Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.
AID
DB 8649
/
6x2c$$$163
08-21-97 06:05:02
dbas
244
Cole and Soelter
REFERENCES
FIG. 10. Genetic evidence for diversity of nuclear divisions in T. thermophila. On the left are illustrations of each of seven types of nuclear division. These are defined (central column), and evidence for their unique genetic control is provided in the right column. Asterisk indicates material from accompanying manuscript on early-stage mutants (Cole et al., 1997).
In higher eukaryotes, one can distinguish four distinct types of nuclear division: meiosis I; meiosis II; the relatively unregulated, rapid nuclear division cycle that accompanies cleavage divisions in early embryogenesis; and the highly regulated cell cycle of more differentiated cell types. Clearly ciliates offer a valuable experimental system for exploring differential regulation of numerous types of nuclear division cycles that have evolved within eukaryotes (see Adl and Berger, 1996, for review). The combination of Tetrahymena’s nuclear dualism with its genetic tractability and easily visualized chromosomal and nuclear dynamics makes it an attractive model system for studying the control of meiosis, cell cycle, and the nuclear behaviors associated with conjugation.
ACKNOWLEDGMENTS The authors acknowledge Phuc Nguyen, Dr. Lauri Sammartano, Jill Hemish, and Amy Wirkkala for hours of Tetrahymena pair isolation. We also thank Kathleen Stuart for her darkroom assistance and Dr. Donna Cassidy-Hanley for her careful criticism of the manuscript. This work was supported by an NSF RUI award (MCB-9303456), an NSF Career Award (MCB-9507285), and an NSF Academic Research Infrastructure Program Award (BIR-9413759).
Adl, S. M., and Berger, J. D. (1994). Cell cycle mutation in Paramecium tetraurelia discriminates between sexual and vegetative functions. Dev. Genet. 15, 172–175. Adl, S. M., and Berger, J. D. (1996). Commitment to division in ciliate cell cycles. J. Eukaryotic Microbiol. 43, 77–86. Allen, S. L. (1967a). Genomic exclusion: A rapid means for inducing homozygous diploid lines in Tetrahymena pyriformis, syngen 1. Science 155, 575–577. Allen, S. L. (1967b). Cytogenetics of genomic exclusion in Tetrahymena. Genetics 55, 797–822. Allen, S. L., File, S. K., and Koch, S. L. (1967). Genomic exclusion in Tetrahymena. Genetics 55, 823–837. Blackburn, E. H. (1991). Structure and function of telomeres. Nature 350, 569–573. Cleffmann, G. (1980). Chromatin elimination and the genetic organisation of the macronucleus in Tetrahymena thermophila. Chromosoma (Berlin) 78, 313–325. Cole, E. S. (1991). Conjugal blocks in Tetrahymena pattern mutants and their cytoplasmic rescue. I. Broadened cortical domains (bcd). Dev. Biol. 148, 403–419. Cole, E. S., and Bruns, P. J. (1992). Uniparental cytogamy: A novel, efficient method for bringing mutations of Tetrahymena into homozygous expression with precocious sexual maturity. Genetics 132, 1017–1031. Cole, E. S., and Frankel, J. (1991). Conjugal blocks in Tetrahymena pattern mutants and their cytoplasmic rescue. II. janus A. Dev. Biol. 148, 420–428. Cole, E. S., Hemish, J., Tuan, G., and Bruns, P. J. (1997). A mutational analysis of conjugation in Tetrahymena thermophila. 1. Phenotypes affecting early development: Meiosis to nuclear selection. Dev. Biol. 189, 215–232. Doerder, F. P. (1979). Regulation of macronuclear DNA content in Tetrahymena thermophila. J. Protozool. 26, 28–35. Doerder, F. P., and DeBault, L. E. (1978). Life cycle variation and regulation of macronuclear DNA content in Tetrahymena thermophila. Chromosoma (Berlin) 69, 1–19. Doerder, F. P., and Shabatura, S. K. (1980). Genomic exclusion in Tetrahymena thermophila: A cytogenetic and cytofluorometric study. Dev. Genet. 1, 205–218. Gaertig, J., and Kaczanowski, A. (1987). Correlation between the shortened period of cell pairing during genomic exclusion and the block in posttransfer nuclear development in Tetrahymena thermophila. Dev. Growth Differ. 29(6), 553–562. Gaertig, J., and Fleury, A. (1992). Spatiotemporal reorganization of intracytoplasmic microtubules is associated with nuclear selection and differentiation during developmental process in the ciliate Tetrahymena thermophila. Protoplasma 167, 74–87. Grandchamp, S., and Beisson, J. (1981). Positional control of nuclear differentiation in Paramecium. Dev. Biol. 81, 336–341. Gray, J. T., Celander, D. W., Price, C. M., and Cech, T. R. (1991). Cloning and expression of genes for the Oxytricha telomere-binding protein: Specific subunit interactions in the telomeric complex. Cell 67, 807–814. Hamilton, E. P. (1984). Dissection of fertilization and development in Tetrahymena using antimicrotubule drugs. [Ph.D. thesis] Univ. of California, Santa Barbara. Hamilton, E. P., Suhr-Jessen, P. B., and Orias, E. (1988). Pronuclear fusion failure: An alternative conjugation pathway in Tetrahy-
Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.
AID
DB 8649
/
6x2c$$$164
08-21-97 06:05:02
dbas
245
Mutational Analysis of Tetrahymena Conjugation, 2
mena thermophila induced by vinblastine. Genetics 118, 627– 636. Kaczanowski, A. (1992). Mutation affecting cell separation and macronuclear resorption during conjugation in Tetrahymena thermophila: Early expression of the zygotic genotype. Dev. Genet. 13, 58–65. Kaczanowski, A., Ramel, M., Kaczanowska, J., and Wheatley, D. (1991). Macronuclear differentiation in conjugating pairs of Tetrahymena treated with the antitubulin drug nocodazole. Exp. Cell Res. 195, 330–337. Kiersnowska, M., Kaczanowski, A., and de Haller, G. (1993). Inhibition of oral morphogenesis during conjugation of Tetrahymena thermophila and its resumption after cell separation. Eur. J. Protistol. 29, 359–369. Nanney, D. L. (1953). Nucleo-cytoplasmic interactions during conjugation in Tetrahymena. Biol. Bull. 105, 133–148. Nelsen, E. M., Williams, N. E., Yi, H., Knaak, J., and Frankel, J. (1994). ‘‘Fenestrin’’ and conjugation in Tetrahymena thermophila. J. Eukaryotic Microbiol. 41(5), 483–495. Orias, E. (1986). Ciliate conjugation. In ‘‘The Molecular Biology of Ciliated Protozoa’’ (L. G. Gall, Ed.), pp. 45–94. Academic Press, Orlando. Orias, E., and Hamilton, E. P. (1979). Cytogamy: An inducible, alternate pathway of conjugation in Tetrahymena thermophila. Genetics 91, 657–671.
Orias, E., Hamilton, E. P., and Flacks, M. (1979). Osmotic shock prevents nuclear exchange and produces whole-genome homozygotes in conjugating Tetrahymena. Science 203, 660–663. Orias, J. D., Hamilton, E. P., and Orias, E. (1983). A microtubular meshwork associated with gametic pronucleus transfer across a cell–cell junction. Science 222, 181–184. Ray, C., Jr. (1956). Meiosis and nuclear behavior in Tetrahymena pyriformis. J. Protozool. 3, 604–610. Rose, D., Thomas, W., and Holm, C. (1990). Segregation of recombined chromosomes in meiosis I requires DNA topoisomerase II. Cell 60, 1009–1017. Sonneborn, T. M. (1954). Patterns of nucleocytoplasmic integration in Paramecium. Carylogia 6, 307–325. Ward, J. G., and Herrick, G. (1996). Effects of transcription inhibitor actinomycin D on postzygotic development of Tetrahymena thermophila conjugants. Dev. Biol. 173, 174–184. Wenkert, D., and Allis, C. D. (1984). Timing of the appearance of micronuclear histone variant hv1 and gene expression in developing new macronuclei of Tetrahymena thermophila. J. Cell Biol. 98, 2107–2117. Yao, M.-C. (1990). Site-specific chromosome breakage and DNA deletion in ciliates. In ‘‘Mobile DNA’’ (D. E. Berg and M. M. Howe, Eds.), pp. 715–734. Am. Soc. Microbiol., Washington, DC. Received for publication March 31, 1997 Accepted June 5, 1997
Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.
AID
DB 8649
/
6x2c$$$164
08-21-97 06:05:02
dbas
A Mutational Analysis of Conjugation in Tetrahymena thermophila 2. Phenotypes Affecting Middle and Late Development: Third Prezygotic Nuclear Division, Pronuclear Exchange, Pronuclear Fusion, and Postzygotic Development
Eric S. Cole*,1 and Timothy A. Soelter* *Biology Department, St. Olaf College, Northfield, Minnesota 55057
Conjugation following pair formation in Tetrahymena can be divided into three distinct sequences of events: prezygotic development, postzygotic development, and exconjugant development. The decision to proceed with postzygotic development is governed by a developmental checkpoint occurring sometime during the middle stages of conjugation. A second developmental decision is made to initiate pair separation and exconjugant development. This paper examines the phenotypes of five newly isolated conjugation mutants (cnj6–cnj10) which affect middle and late events within the conjugation program. cnj6 mutants exhibit normal nuclear behavior throughout development up to and including differentiation of new macronuclear anlagen. Pairs arrest at this developmental endpoint, unable to dissociate. cnj7 and cnj8 eliminate the third prezygotic nuclear division and the first postzygotic nuclear division. All subsequent developmental events appear normal. cnj9 eliminates the second postzygotic nuclear division, and subsequently, new macronuclei fail to develop despite parental macronuclear degradation. cnj10 results in a pleiotropic phenotype characterized by failure of numerous events which all appear to involve nuclear–cytoskeletal interactions. These defects include nuclear selection (anchoring nuclei to the exchange junction), pronuclear exchange, pronuclear fusion, and anchoring postzygotic nuclear division products to the posterior cell cortex. These mutant phenotypes are used to draw inferences regarding developmental dependencies that govern a cell’s entry into the postzygotic and exconjugant developmental programs. q 1997 Academic Press
INTRODUCTION One of the more intriguing aspects of conjugation in Tetrahymena is the functional separation of prezygotic, postzygotic, and ‘‘exconjugant’’ developmental subprograms. Prezygotic development is initiated by pair formation and includes two meiotic nuclear divisions, selection of one meiotic product to undergo a third nuclear mitosis, differentiation of a migratory and stationary pronucleus within each mating partner, exchange of migratory pronuclei, and fusion of the migratory pronucleus with the resident, stationary pronucleus, resulting in formation of a zygotic nucleus. The earliest events within this program, meiosis and nuclear ‘‘selection,’’ are described in the accompanying paper (Cole 1
To whom correspondence should be addressed. Fax: (507) 6463968. E-mail: [email protected].
et al., 1997). Although nuclear fusion formally marks the beginning of postzygotic development, it is unclear what event or conditions actually trigger this transition. What is clear is that when the appropriate conditions are not met, cells abort development, retain their parental macronuclei, and separate prematurely in a process known as genomic exclusion (Allen, 1967a,b; Allen et al., 1967; Doerder and Shabatura, 1980). Postzygotic events involve two consecutive nuclear divisions. The first postzygotic division spindle delivers nuclei into the anterior and posterior cytoplasm. The anterior nucleus migrates more posteriorly prior to the second division (Ray, 1956). The spindle of the second division delivers two of four nuclear products to the anterior cytoplasm where a program of chromosomal modifications is initiated transforming these two nuclei into macronuclear anlagen (MA; see Nanney, 1953). The two remaining nuclei are positioned in the posterior cytoplasm
0012-1606/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.
AID
DB 8649
/
6x2c$$$161
233
08-21-97 06:05:02
dbas
234
Cole and Soelter
where they appear to become ‘‘anchored’’ to the posterior cell cortex (Kaczanowski et al., 1991). These posterior nuclei differentiate into germ-line micronuclei. Macronuclear anlagen formation is the second case in which the anterior (paroral) cytoplasm appears to exert a determinative influence over nuclear fate. During prezygotic development, the anteriormost, paroral nucleus is ‘‘selected’’ to undergo a third, prezygotic division (see accompanying manuscript, Cole et al., 1997). In Paramecium, the situation is different in that the oral cavity is situated in the posterior third of the cell. Perhaps for this reason, it is the posterior nuclear products which differentiate into macronuclear anlagen within this species (Sonneborn, 1954). In both species, differentiation of somatic and germinal nuclei is associated with extreme polar localization, a phenomenon reminiscent of pole cell nuclear determination in the embryogenesis of Drosophila (as pointed out by Grandchamp and Beisson, 1981). It has been suggested that any nuclei migrating into the anterior cytoplasm (in Tetrahymena) will differentiate as macronuclei, but anchoring of micronuclei at the cell’s posterior cortex is necessary to protect them from macronuclear differentiation and preserve their germinal, micronuclear character (Kaczanowski et al., 1991). Macronuclear differentiation involves chromosome fragmentation, DNA elimination, telomere synthesis, and gene amplification (Yao, 1990; Blackburn, 1991; Gray et al., 1991). Eight to ten hours after cell mixing, transcription is detected in the newly formed macronuclear anlagen (Wenkert and Allis, 1984). Twelve hours after mixing (at 307C), if all nuclear events occur normally, pairs dissociate, the parental macronucleus is destroyed, and one of the two micronuclei is eliminated. Micronuclear elimination and macronuclear resorption are accompanied by regeneration of the oral apparatus (oral replacement) at approximately 14 hr (Cole and Frankel, 1991; Kiersnowska and Kaczanowski, 1993). These events will be referred to here as ‘‘exconjugant’’ development. This investigation examined five novel conjugation mutants, cnj6–cnj10, which, considered with two previously characterized mutants, bcd and janA (Cole, 1991; Cole and Frankel, 1991), affect each of the middle and late stages of nuclear behavior. In particular we have used the mutant phenotypes to shed light on developmental decisions which govern the transition from prezygotic to postzygotic and from postzygotic to exconjugant development.
METHODS AND MATERIALS The methods are described in the accompanying paper (Cole et al., 1997).
RESULTS Middle and Late Stages of Conjugation in Tetrahymena thermophila Figure 1 shows a schematic diagram of Tetrahymena conjugation. Figure 2 shows DAPI-stained representatives of
the developmental events beginning with the gametogenic (third) nuclear division (Figs. 2A and 2B) and ending with the three exconjugant stages (Figs. 2M, 2N, and 2O).
Phenotypic Profiles of Conjugation Mutants cnj7 and cnj8. cnj7 and cnj8 were originally isolated based upon quite similar terminal pair configurations with abnormal numbers of macronuclear anlagen and micronuclei. The most frequent endpoint consisted of a single, enlarged macronuclear anlagen and a single, enlarged micronucleus (Fig. 3B). Developmental analysis revealed that the earliest abnormality appeared after the completion of meiosis II (Fig. 4). At this point, the third prezygotic (gametogenic) division failed to occur. Normally, we would predict that the four haploid products would subsequently degenerate, the cells would become ‘‘starlike,’’ and development would abort following the genomic exclusion pathway (Allen, 1967a,b; Allen et al., 1967; Doerder and Shabatura, 1980). However, in cnj7 and cnj8, one of the four meiotic nuclei became associated with the exchange junction (Fig. 3A) and behaved as a migratory pronucleus while a second meiotic product appeared to differentiate as a stationary pronucleus. Both of these nuclei were decorated with fenestrin (Figs. 5A and 5B), a protein associated with differentiated pronuclei (Nelsen et al., 1994). Migratory pronuclei appeared to be exchanged, and synkarya were observed, indicating that pronuclear fusion (karyogamy) had occurred. Subsequent to nuclear fusion, there was a frequent failure of one of the two postzygotic mitoses. The most abundant endpoint possessed both a single micronucleus and a single MA, suggesting that the first division had failed and the second postzygotic division had occurred, thereby delivering one nucleus to the anterior cytoplasm (where it differentiated into a macronuclear anlagen) and one nucleus to the posterior cytoplasm where it became anchored and differentiated as a germinal micronucleus. Our temporal profile of (cnj7 1 cnj7) was compared to the wild-type profile shown in the accompanying paper (Cole et al., 1997) to distinguish whether the skipped nuclear division was the first or second postzygotic division. (Note. The cnj8 profile was very similar to the cnj7 profile and hence is not shown.) Exact timing of developmental events varies from experiment to experiment. Hence, a good internal marker is the developmental appearance of the pronuclear exchange configuration (Fig. 1J). From the wild-type profile, we saw that pairs which completed the first postzygotic division appeared 12 hr following the appearance of pairs in the exchange configuration, whereas second postzygotic division figures appeared 1 hr after the exchange configuration. From Fig. 5, we see that cnj7 pairs exhibiting the exchange configuration first appeared 4.5 – 5 hr after cell mixing. Pairs exhibiting two postzygotic nuclei appeared an hour or more after first appearance of ex-
Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.
AID
DB 8649
/
6x2c$$$162
08-21-97 06:05:02
dbas
235
Mutational Analysis of Tetrahymena Conjugation, 2
FIG. 1. A schematic diagram showing all of conjugal development in Tetrahymena thermophila.
change configurations. These results strongly suggest that cnj7 (and cnj8) pairs skip the first postzygotic division and resume development with the second postzygotic division. A heterotypic mating between (cnj7 1 cnj8) revealed complementation and a fairly high degree of cytoplasmic rescue of the conjugal phenotype. This strongly suggests that these mutations, despite their similar phenotypes, do in fact define distinct loci. cnj9. cnj9 matings exhibited perfectly normal chromosomal and nuclear behavior up to the second postzygotic mitosis (Fig. 6). This final nuclear division apparently failed in this phenotype with intriguing consequences. The parental macronucleus condensed on schedule, and yet, neither of the two postzygotic nuclei differentiated into macronuclear anlagen (see Figs. 3C and 3D). Consequently, the most common endpoint for cnj9 was a pair of cells, each with a condensed parental macronucleus and two (enlarged) micronuclei. It should be added that penetrance was not 100%. Approximately 5% of pairs showed macronuclear anlagen formation (Fig. 3D), and it is notable that in these rare cases, the second postzygotic division was also successful. (We
have found very rare cases in which only two zygotic nuclei were present and one had differentiated into a macronuclear anlagen.) Again, developmental timing convinced us that the first postzygotic nuclear division occurred on schedule, whereas the second postzygotic division failed. In (cnj9 1 cnj9) matings, pairs exhibiting two postzygotic nuclei appeared just 12 hr after first appearance of exchange configurations (Fig. 6). cnj10. The cnj10 mutant was initially isolated based upon a terminal phenotype in which pairs contained multiple macronuclear anlagen. Cells with as many as eight nuclear figures were observed (Fig. 3F). Cells homozygous for the cnj10 mutation paired normally and appeared to undergo normal development up to the third prezygotic division (see developmental profile, Fig. 7). At this point, there was a high incidence of pairs in which at least one partner eliminated all four meiotic products. In effect, nuclear selection failed quite frequently. Analysis of asymmetric pairs in which one partner exhibited nuclear selection and the other eliminated all four meiotic products was quite revealing. In such pairs it was possible to observe whether the
Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.
AID
DB 8649
/
6x2c$$$162
08-21-97 06:05:02
dbas
236
Cole and Soelter
08-21-97 06:05:02
dbas
237
Mutational Analysis of Tetrahymena Conjugation, 2
FIG. 3. (A and B) cnj7 pairs. (A) The exchange configuration for a cnj7 pair. Note total of four nuclei per cell, one at exchange junction. (B) Typical cnj7 endpoint. (Note single enlarged MA and single enlarged micronucleus, Mic). (C and D) cnj9 1 cnj9 pairs: abnormal endpoints. (C) Most typical cnj9 endpoint showing condensed parental macronucleus (PM) and two enlarged micronuclei. (D) Rare cnj9 endpoint with one normal partner. (E and F) cnj10 pairs. (E) A cnj10 pair showing the second postzygotic division spindles (note four spindles will generate eight nuclear products). Also note the unilateral development characteristic of nuclear selection failure in left-hand partner earlier in development. (F) Typical cnj10 endpoint with eight nuclei in one partner (many or most of them developing as macronuclear anlagen).
normally developing partner could transfer its migratory pronucleus unilaterally to the defective partner. In fact, this was never observed. Cells which completed the third prezygotic division were never observed to adopt the ‘‘exchange configuration’’ (see Fig. 2C). Instead, pronuclei maintained their roundness and their distance from the exchange junction. Subsequently, even in pairs in which both partners developed migratory and stationary pronuclei, synkaryon formation was never observed and the two pronuclei per-
sisted as distinct entities. The three ‘‘relic’’ nuclei were eliminated normally. In cell partners which successfully completed the third prezygotic nuclear division, the two persistent pronuclei underwent two postzygotic divisions, resulting in as many as eight mitotic products (Fig. 3F). Curiously, all of these nuclei tended to aggregate in the cytoplasm anterior to the macronucleus. Consequently, most of these nuclei initiated macronuclear anlagen development (Fig. 3F). One final observation regarding cnj10 was
FIG. 2. DAPI-stained fluorescence micrographs of middle and late stages in normal conjugal development. In each panel, nuclei from a symmetrical pair of mating cells are shown. (A) Third prezygotic division showing anaphase chromosomes (open arrow) and secondary ‘‘pseudospindle’’ (closed arrow). Densely staining small round bodies are ‘‘nuclear relics,’’ i.e., nuclei which have been targeted for elimination. (B) Completion of third prezygotic division showing still condensed mitotic chromosomes. (C) Pronuclei have assumed exchange configuration (M, migratory and S, stationary pronuclei). Note decondensed nature of pronuclei. (D) After pronuclear exchange, pronuclei have fused (arrow) forming synkaryon. (E and F) First postzygotic nuclear division (PZD). (G) Completion of first PZD. Note two rounded nuclear products. (H and I) Second PZD. (J) Completion of second PZD. Note that the anterior pair of nuclei have already assumed slightly larger dimensions. (K and L) Macronuclear anlagen differentiation. Closed arrows indicate differentiating MA. Open arrow indicates condensing parental macronucleus. (M) Pair separation. Note two enlarged macronuclear anlagen (diffuse), two brightly staining micronuclei, and the condensing parental macronucleus. (N) Parental macronucleus has been eliminated. (O) Second micronucleus has been eliminated.
Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.
AID
DB 8649
/
6x2c$$$162
08-21-97 06:05:02
dbas
238
Cole and Soelter
FIG. 4. Developmental profile for cnj7 1 cnj7. 100 pairs were scored for each time point. The y axis represents the percentage of the sample in given stages for each given time point.
that postzygotic developmental divisions were somewhat delayed. cnj6 and janA. Finally we have two mutations that affect very late stages in development. Both janA (see Cole and Frankel, 1991) and cnj6 result in a failure of pair separation after successful completion of a normal sequence of nuclear events (see also the mra mutation; Kaczanowski, 1992). These mutant cell lines apparently fail to produce some essential trigger for pair termination. In these pairs, one of the two micronuclei and the condensed, parental macronucleus failed to be resorbed at the appropriate time.
shown in Fig. 8. These phenotypes provide us with a unique opportunity to explore the dependencies and checkpoints which govern this elaborate developmental program. In particular, these mutants offer insights regarding the decision to initiate or trigger postzygotic development, the decision to initiate macronuclear anlagen formation, and the decision for pairs to separate and initiate exconjugant development (see Fig. 9 for a summary of middle and late stage developmental dependencies).
Developmental Dependencies
DISCUSSION Overview A summary of mutant phenotypes affecting middle and late stages of conjugal development in Tetrahymena is
Triggering postzygotic development. Until pronuclei fuse forming the synkaryon, mating partners are not committed to pursue the postzygotic developmental program which includes the first and second postzygotic nuclear di-
Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.
AID
DB 8649
/
6x2c$$$162
08-21-97 06:05:02
dbas
239
Mutational Analysis of Tetrahymena Conjugation, 2
FIG. 5. A cnj7 1 cnj7 mutant pair stained with both (A) 3A7, a monoclonal antiserum directed against the protein fenestrin visualized with an FITC-conjugated secondary antiserum, and (B) DAPI nuclear stain. Note fenestrin labeling of two (meiotic) nuclear products.
visions and macronuclear anlagen differentiation. In fact, pairs that run into difficulties within the early stages of development frequently retain their parental macronuclei and abort subsequent development (a process termed firstround genomic exclusion; Allen, 1967a,b; Doerder and Shabatura, 1980). There are four conspicuous candidates for the cytological trigger that commits a cell to postzygotic development: (1) completion of the third prezygotic division, (2) association of two sets of migratory pronuclei with the exchange junction, (3) pronuclear exchange, and (4) pronuclear fusion or karyogamy. Here we summarize data which suggest that none of these events is sufficient to trigger postzygotic development. Is the third prezygotic division necessary in order to trigger postzygotic development? The evidence suggests not. The third prezygotic division appears to be eliminated in both the cnj7 and the cnj8 mutants, yet postzygotic development proceeds relatively normally. Hence, the third prezygotic division is not necessary in order to trigger postzygotic development. Association of the migratory pronucleus with the exchange junction has been postulated as the event that triggers postzygotic development (Hamilton, 1984). This must happen in both partners as evidenced by genomic exclusion matings in which one partner develops normally, while the other partner eliminates all of its meiotic nuclei (Allen, 1967a,b; Doerder and Shabatura, 1980). The result is unilateral pronuclear development and transfer and abortive development; postzygotic development is not initiated. cnj10 pairs defy this hypothesis as do previously described bcd matings (Cole, 1991). cnj10 matings frequently become ‘‘unilateral’’ in that one partner destroys all of its meiotic nuclei, yet development proceeds in the partner which has successfully undergone nuclear
selection and third-prezygotic division. In fact, development proceeds (unilaterally) all the way through macronuclear anlagen formation. Hence, bilateral pronuclear association with the exchange junction is not necessary for triggering postzygotic development. This is also demonstrated by another type of aberrant mating. Uniparental cytogamy is a form of genomic exclusion mating in which self-fertilization is provoked by osmotic shock (Cole and Bruns, 1992). In these pairs, unilateral pronuclear association with the exchange junction occurs, yet cells initiate self-fertilization and postzygotic development in the normal partner. Finally, bcd matings demonstrate that bilateral pronuclear association with the exchange junction is insufficient to trigger postzygotic development. In these cells, migratory pronuclei are successfully exchanged between mating partners, yet they fail to fuse and pairs abort development (Cole, 1991). Hence, it would appear that bilateral association of pronuclei with the exchange junction is neither necessary nor sufficient to trigger postzygotic development. Pronuclear exchange can be blocked in a number of ways. Vinblastine and hyperosmotic shocks have both been shown to prevent pronuclear exchange (Hamilton and SuhrJessen, 1980; Orias et al., 1979; Orias and Hamilton, 1979), yet postzygotic development proceeds. Again, cnj10 cells also exhibit a failure of pronuclear transfer and yet development proceeds. Pronuclear fusion (karyogamy) can also be blocked using anti-microtubule drugs such as vinblastine (Hamilton, 1984; Hamilton et al., 1988) and nocodazole (Kaczanowski et al., 1991) and yet macronuclear anlagen formation and parental macronucleus condensation occur (both hallmarks of successful postzygotic development). Our cnj10 phenotype supports these findings in that cnj10 pairs exhibit pronuclear fusion failure and yet proceed with postzygotic de-
Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.
AID
DB 8649
/
6x2c$$$163
08-21-97 06:05:02
dbas
240
Cole and Soelter
FIG. 6. Developmental profile for cnj9 1 cnj9. 100 pairs were scored for each time point. The y axis represents the percentage of the sample in given stages for each given time point.
velopment. Hence, pronuclear fusion is not necessary in order to provoke postzygotic development. All these observations serve to illustrate that, so far, there is no simple, conspicuous cytological event that can be identified as the trigger for postzygotic development. This suggests that such an important developmental decision may be controlled by more than one mechanism and that it is developmentally buffered. This would make adaptive sense in that commitment to postzygotic development is also a commitment to destroying one’s functional somatic macronucleus. Further insight into the control of this developmental decision has been provided by Ward and Herrick (1996). They have shown that, whatever the cytological trigger may be, it must activate novel gene transcription and translation in order to provoke entry into postzygotic development. Triggering macronuclear anlagen formation. It has
already been well established that in Tetrahymena, micronuclei must be translocated to the anterior cytoplasm in order for them to begin differentiating into macronuclear anlagen (Nanney, 1953). Our cnj9 mutant is of interest in this regard. It is the second postzygotic division that delivers two micronuclei to the anterior cytoplasm and two to the posterior cytoplasm, and it is this same division which appears to be blocked in cnj9 matings. This is consistent with our observations that cnj9 cells also fail to differentiate macronuclear anlagen. Orias reports that vinblastine treatment during this stage results (essentially) in a phenocopy of our cnj9 mutation. He writes: ‘‘When (and only when) the last (i.e., second) postzygotic nuclear division is inhibited, macronuclear differentiation is delayed, and nuclear differentiation products intermediate in appearance between micro- and macro-nucleus are often observed (staining almost as
Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.
AID
DB 8649
/
6x2c$$$163
08-21-97 06:05:02
dbas
241
Mutational Analysis of Tetrahymena Conjugation, 2
FIG. 7. Developmental profile for cnj10 1 cnj10. 100 pairs were scored for each time point. The y axis represents the percentage of the sample in given stages for each given time point.
densely as normal new micronuclei, but larger than them in size and apparent amount of DNA)’’ (Orias, 1986, p. 69). This is a perfect description of the cnj9 phenotype. One other piece of information that our cnj9 mutant provides is that condensation of the parental macronucleus is not dependent upon macronuclear anlagen formation. Macronuclear condensation must therefore be triggered by gene products produced by this very same parental macronucleus. Similar observations have been reported by others (Nanney, 1953; Doerder and Shabatura, 1980). Macronuclear condensation is a classic example of developmentally programmed nuclear elimination. The cnj10 phenotype is also informative. The cnj10 mutation appears remarkably pleiotropic. The earliest cnj10 defect appears to be failure of nuclear selection (a prezygotic activity). A second defect appears to be an inability of the cell to transfer its migratory pronucleus
across the exchange junction. Subsequent defects include pronuclear fusion failure, postzygotic developmental delay (at least in asymmetric pairs with one starlike partner, which resemble the developmental delay reported by Gaertig and Kaczanowski, 1987), and a consequent production of supernumerary macronuclear anlagen. The supernumerary MA can be understood as the consequence of two events. First, without nuclear fusion, there are two nuclei which undergo both postzygotic divisions. This can result in as many as eight postzygotic nuclear division products. The second defect appears to be a failure of cnj10 cells to anchor the posterior nuclear division products at the posterior cortex. Kaczanowski has argued that without this cortical anchoring, nuclei migrate anteriorly and differentiate as MA (Kaczanowski et al., 1991). It is notable that at least three events involving intimate nuclear – cortical association appear defective in cnj10: se-
Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.
AID
DB 8649
/
6x2c$$$163
08-21-97 06:05:02
dbas
242
Cole and Soelter
FIG. 8. Summary of nuclear behavior during middle and late stages of conjugation and phenotypes of some conjugation mutants. Lettering refers to the stages depicted in Fig. 1.
lection of one of the four meiotic products within the protective paroral cortex, pronuclear migration across the intercellular exchange junction, and anchoring of the posterior products of the second postzygotic division to the posteriormost cortex to maintain their germinal, micronuclear character. These defects argue that cnj10 plays a role in nuclear – cytoskeletal association, thereby supporting the hypothesis that cortical anchoring protects a nucleus from various cytoplasmic signals of differentiation. Triggering pair separation and exconjugant development. Reagents that block transcription or translation, when delivered after macronuclear anlagen formation has begun, result in a pair-separation-failure syndrome (Ward and Herrick, 1996), which phenocopies the cnj6 mutant as well as janA and mrA (Cole and Frankel, 1991; Kaczanowski, 1992). This condition includes an inability of pairs to separate, failure of pairs to eliminate one of their two micronuclei, and an inability of pairs to eliminate their old condensed parental macronucleus. This condition is lethal. These results suggest that there is a sensitive developmental stage during which pairs become competent to com-
plete late conjugal and postconjugal development. It also suggests that, whereas most of development is under the genetic control of the parental macronucleus, pair separation and exconjugant development are under the genetic control of the newly formed macronuclear anlagen. We propose that these mutations (for various reasons) are unable to effect the transition from parental to zygotic nuclear gene expression. This hypothesis is currently under investigation.
Diversity of Nuclear Divisions From our studies of all 10 cnj mutants (see accompanying paper, Cole et al., 1997), the case can be made that Tetrahymena carry out seven distinct types of nuclear division (see Fig. 10). During vegetative cell growth, the micronucleus and macronucleus exhibit profoundly different types of nuclear division. The macronucleus undergoes an amitotic fission which can be functionally dissociated from the cycle of DNA synthesis (Cleffmann, 1980; Doerder and DeBault, 1978; Doerder, 1979). The micronucleus undergoes a more conventional mitosis during vegetative growth, albeit still
Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.
AID
DB 8649
/
6x2c$$$163
08-21-97 06:05:02
dbas
243
Mutational Analysis of Tetrahymena Conjugation, 2
FIG. 9. Developmental dependencies during middle and late, postzygotic conjugation in Tetrahymena. Bold arrows indicate inferred developmental dependencies. For example, pronuclear differentiation appears to be required for pronuclear exchange and pronuclear fusion. However, pronuclear fusion does not depend upon exchange (hence no arrow). Mutant designations in italics refer to those genes whose wild-type activities appear necessary for a given step. Trigger to postzygotic development, indicated by the vertical box, depends upon some as yet unidentified cytological event(s), and appears necessary to trigger an ensemble of events including the two postzygotic nuclear divisions, macronuclear anlagen differentiation, and parental macronucleus condensation. Each of these four events appears to be able to take place independently from the other three, hence they each have individual arrows driving them. It is unknown whether transcriptional activation of the macronuclear anlagen depends upon the chromosomal rearrangements that characterize macronuclear differentiation (hence the question mark). Transcriptional activation does appear to be essential for all the events associated with ‘‘exconjugant development’’ (listed under Pair Separation). Parental macronucleus condensation appears to occur independently of pair separation (hence no arrow connecting them).
within a closed nuclear envelope. During the early stages of conjugation the micronucleus undergoes three distinct types of karyokinesis: meiosis I (which involves synapsis, recombination, and absence of centromere division), meiosis II (which does exhibit centromere division), and a third ‘‘gametogenic’’ mitosis directly following a round of DNA synthesis (Doerder and Shabatura, 1980). That meiosis I and II are distinct comes as no surprise, but it is intriguing to learn that all three prezygotic divisions are distinct from one another and from the vegetative mitoses. The case for this can be made as follows. During conjugation, wild-type cnj1 and cnj2 genes are clearly essential for all three prezygotic nuclear divisions (see accompanying paper by Cole et al., 1997). Specifically, cells homozygous for the cnj1 or cnj2 mutations appear unable execute chromosome condensation; chromosomes fail to segregate, and consequently nuclei fail to divide. It would appear that cnj1 and cnj2 gene products are essential for all three prezygotic divisions. Nevertheless, the vegetative micronuclear mitosis occurs freely in cnj1 and cnj2 homozy-
gotes exhibiting normal chromatin condensation. Hence, in this way at least, vegetative micronuclear mitoses are distinct from meiosis I, meiosis II, and the third prezygotic mitosis. A complementary situation has been observed in Paramecium tetraurelia. Adl and Berger (1994) have shown that the temperature-sensitive cc1 mutation blocks micronuclear elongation (and oral development) during the vegetative cell cycle, but exhibits normal micronuclear divisions (and oral replacement) during conjugal development. The fact that cnj5 pairs skip both meiotic divisions and yet proceed through the third (gametogenic) mitosis distinguishes this third mitosis from the second meiotic division, even though both superficially resemble one another (see accompanying paper, Cole et al., 1997). The cnj7 and cnj8 mutations both appear to cause elimination of the first postzygotic mitosis but not the second. The cnj9 mutation clearly eliminates the second postzygotic mitosis but not the first. These results argue that each of these seven different nuclear divisions requires the activity of different gene products.
Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.
AID
DB 8649
/
6x2c$$$163
08-21-97 06:05:02
dbas
244
Cole and Soelter
REFERENCES
FIG. 10. Genetic evidence for diversity of nuclear divisions in T. thermophila. On the left are illustrations of each of seven types of nuclear division. These are defined (central column), and evidence for their unique genetic control is provided in the right column. Asterisk indicates material from accompanying manuscript on early-stage mutants (Cole et al., 1997).
In higher eukaryotes, one can distinguish four distinct types of nuclear division: meiosis I; meiosis II; the relatively unregulated, rapid nuclear division cycle that accompanies cleavage divisions in early embryogenesis; and the highly regulated cell cycle of more differentiated cell types. Clearly ciliates offer a valuable experimental system for exploring differential regulation of numerous types of nuclear division cycles that have evolved within eukaryotes (see Adl and Berger, 1996, for review). The combination of Tetrahymena’s nuclear dualism with its genetic tractability and easily visualized chromosomal and nuclear dynamics makes it an attractive model system for studying the control of meiosis, cell cycle, and the nuclear behaviors associated with conjugation.
ACKNOWLEDGMENTS The authors acknowledge Phuc Nguyen, Dr. Lauri Sammartano, Jill Hemish, and Amy Wirkkala for hours of Tetrahymena pair isolation. We also thank Kathleen Stuart for her darkroom assistance and Dr. Donna Cassidy-Hanley for her careful criticism of the manuscript. This work was supported by an NSF RUI award (MCB-9303456), an NSF Career Award (MCB-9507285), and an NSF Academic Research Infrastructure Program Award (BIR-9413759).
Adl, S. M., and Berger, J. D. (1994). Cell cycle mutation in Paramecium tetraurelia discriminates between sexual and vegetative functions. Dev. Genet. 15, 172–175. Adl, S. M., and Berger, J. D. (1996). Commitment to division in ciliate cell cycles. J. Eukaryotic Microbiol. 43, 77–86. Allen, S. L. (1967a). Genomic exclusion: A rapid means for inducing homozygous diploid lines in Tetrahymena pyriformis, syngen 1. Science 155, 575–577. Allen, S. L. (1967b). Cytogenetics of genomic exclusion in Tetrahymena. Genetics 55, 797–822. Allen, S. L., File, S. K., and Koch, S. L. (1967). Genomic exclusion in Tetrahymena. Genetics 55, 823–837. Blackburn, E. H. (1991). Structure and function of telomeres. Nature 350, 569–573. Cleffmann, G. (1980). Chromatin elimination and the genetic organisation of the macronucleus in Tetrahymena thermophila. Chromosoma (Berlin) 78, 313–325. Cole, E. S. (1991). Conjugal blocks in Tetrahymena pattern mutants and their cytoplasmic rescue. I. Broadened cortical domains (bcd). Dev. Biol. 148, 403–419. Cole, E. S., and Bruns, P. J. (1992). Uniparental cytogamy: A novel, efficient method for bringing mutations of Tetrahymena into homozygous expression with precocious sexual maturity. Genetics 132, 1017–1031. Cole, E. S., and Frankel, J. (1991). Conjugal blocks in Tetrahymena pattern mutants and their cytoplasmic rescue. II. janus A. Dev. Biol. 148, 420–428. Cole, E. S., Hemish, J., Tuan, G., and Bruns, P. J. (1997). A mutational analysis of conjugation in Tetrahymena thermophila. 1. Phenotypes affecting early development: Meiosis to nuclear selection. Dev. Biol. 189, 215–232. Doerder, F. P. (1979). Regulation of macronuclear DNA content in Tetrahymena thermophila. J. Protozool. 26, 28–35. Doerder, F. P., and DeBault, L. E. (1978). Life cycle variation and regulation of macronuclear DNA content in Tetrahymena thermophila. Chromosoma (Berlin) 69, 1–19. Doerder, F. P., and Shabatura, S. K. (1980). Genomic exclusion in Tetrahymena thermophila: A cytogenetic and cytofluorometric study. Dev. Genet. 1, 205–218. Gaertig, J., and Kaczanowski, A. (1987). Correlation between the shortened period of cell pairing during genomic exclusion and the block in posttransfer nuclear development in Tetrahymena thermophila. Dev. Growth Differ. 29(6), 553–562. Gaertig, J., and Fleury, A. (1992). Spatiotemporal reorganization of intracytoplasmic microtubules is associated with nuclear selection and differentiation during developmental process in the ciliate Tetrahymena thermophila. Protoplasma 167, 74–87. Grandchamp, S., and Beisson, J. (1981). Positional control of nuclear differentiation in Paramecium. Dev. Biol. 81, 336–341. Gray, J. T., Celander, D. W., Price, C. M., and Cech, T. R. (1991). Cloning and expression of genes for the Oxytricha telomere-binding protein: Specific subunit interactions in the telomeric complex. Cell 67, 807–814. Hamilton, E. P. (1984). Dissection of fertilization and development in Tetrahymena using antimicrotubule drugs. [Ph.D. thesis] Univ. of California, Santa Barbara. Hamilton, E. P., Suhr-Jessen, P. B., and Orias, E. (1988). Pronuclear fusion failure: An alternative conjugation pathway in Tetrahy-
Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.
AID
DB 8649
/
6x2c$$$164
08-21-97 06:05:02
dbas
245
Mutational Analysis of Tetrahymena Conjugation, 2
mena thermophila induced by vinblastine. Genetics 118, 627– 636. Kaczanowski, A. (1992). Mutation affecting cell separation and macronuclear resorption during conjugation in Tetrahymena thermophila: Early expression of the zygotic genotype. Dev. Genet. 13, 58–65. Kaczanowski, A., Ramel, M., Kaczanowska, J., and Wheatley, D. (1991). Macronuclear differentiation in conjugating pairs of Tetrahymena treated with the antitubulin drug nocodazole. Exp. Cell Res. 195, 330–337. Kiersnowska, M., Kaczanowski, A., and de Haller, G. (1993). Inhibition of oral morphogenesis during conjugation of Tetrahymena thermophila and its resumption after cell separation. Eur. J. Protistol. 29, 359–369. Nanney, D. L. (1953). Nucleo-cytoplasmic interactions during conjugation in Tetrahymena. Biol. Bull. 105, 133–148. Nelsen, E. M., Williams, N. E., Yi, H., Knaak, J., and Frankel, J. (1994). ‘‘Fenestrin’’ and conjugation in Tetrahymena thermophila. J. Eukaryotic Microbiol. 41(5), 483–495. Orias, E. (1986). Ciliate conjugation. In ‘‘The Molecular Biology of Ciliated Protozoa’’ (L. G. Gall, Ed.), pp. 45–94. Academic Press, Orlando. Orias, E., and Hamilton, E. P. (1979). Cytogamy: An inducible, alternate pathway of conjugation in Tetrahymena thermophila. Genetics 91, 657–671.
Orias, E., Hamilton, E. P., and Flacks, M. (1979). Osmotic shock prevents nuclear exchange and produces whole-genome homozygotes in conjugating Tetrahymena. Science 203, 660–663. Orias, J. D., Hamilton, E. P., and Orias, E. (1983). A microtubular meshwork associated with gametic pronucleus transfer across a cell–cell junction. Science 222, 181–184. Ray, C., Jr. (1956). Meiosis and nuclear behavior in Tetrahymena pyriformis. J. Protozool. 3, 604–610. Rose, D., Thomas, W., and Holm, C. (1990). Segregation of recombined chromosomes in meiosis I requires DNA topoisomerase II. Cell 60, 1009–1017. Sonneborn, T. M. (1954). Patterns of nucleocytoplasmic integration in Paramecium. Carylogia 6, 307–325. Ward, J. G., and Herrick, G. (1996). Effects of transcription inhibitor actinomycin D on postzygotic development of Tetrahymena thermophila conjugants. Dev. Biol. 173, 174–184. Wenkert, D., and Allis, C. D. (1984). Timing of the appearance of micronuclear histone variant hv1 and gene expression in developing new macronuclei of Tetrahymena thermophila. J. Cell Biol. 98, 2107–2117. Yao, M.-C. (1990). Site-specific chromosome breakage and DNA deletion in ciliates. In ‘‘Mobile DNA’’ (D. E. Berg and M. M. Howe, Eds.), pp. 715–734. Am. Soc. Microbiol., Washington, DC. Received for publication March 31, 1997 Accepted June 5, 1997
Copyright q 1997 by Academic Press. All rights of reproduction in any form reserved.
AID
DB 8649
/
6x2c$$$164
08-21-97 06:05:02
dbas