Molecular and genetic studies on regulation of the cell cycle in Paramecium and other ciliates

J.D Berger laboratory
Department of Zoology
University of British Columbia



Topic 1. Regulation of cyclin-dependent kinases and cyclins

Topic 2. Molecular Analysis of the cc1 rescuing factor

Topic 3 Cell cycle regulation in other ciliates.

Summary of proposal

Most of the work analyzing cell cycle control has been done in higher eukaryotes (animals or plants). Relatively little is known about the regulation of the cell cycle in lower eukaryotes (algae, protozoa). The object of this research is to define and characterize at the molecular and genetic level factors involved in regulation of the cell cycle in Paramecium and other ciliates (Protozoa). This work builds on extensive earlier work characterizing the cell cycle at the physiological level. Two approaches have been used in pursuing analysis at the molecular level. First, a number of components have been identified on the basis of homology with factors involved in regulation of the cell division cycle in higher eukaryotes. We have cloned and partially characterized two classes of cyclin-dependent kinases (cdks) and two classes of cyclins in Paramecium. These are the major regulatory molecules of the cell cycle. A major focus of the proposal is analysis of their functions during the cell cycle through assay of protein kinase activity and subcellular localization. A second approach has been to identify and analyse novel factors involved in cell cycle regulation in Paramecium. We identified a gene locus (cc1) that is required for cell cycle progression in G1, for initiation and maintenance of macronuclear DNA synthesis and for commitment to division. In the last year we obtained a 1.8kb DNA sequence that rescues the mutant phenotype when injected into macronuclei. This sequence contains a novel coding region, that appears to be well conserved within ciliates, based on our initial PCR studies. We hope that analysis of its function and its relationship with cdks and cyclins will provide insights into some of the unique features of ciliate cell cycle regulation. As we have identified homologous sequences for cyclins, cdks and the cc1 rescuing sequence in Tetrahymena, we propose to carry out a comparative analysis of cell cycle regulatory functions in these two ciliates


Our objective has been to explore regulation of the cell cycle in Paramecium and other ciliates through a combination of molecular and genetic approaches. In particular, we have made substantial headway in analysis of cdk (cyclin-dependent kinase) proteins and their associated cyclins in Paramecium. We propose to continue this work and to initiate molecular analysis of the cc1 gene. The cc1ts mutation has a unique cell cycle phenotype (explained below, Topic 2) and has been extensively analyzed in this laboratory at the physiological and genetic level [1]. We will also extend comparative molecular analyses of cell cycle control to other ciliates Tetrahymena and Sterkiella.

Background - Paramecium cell cycle. We have used two approaches in analysis of cell cycle function. A) Exploration of the shared cell cycle regulatory machinery that is the heritage of all eukaryotes (cdks, cyclins etc.), starting with homologies to other systems, and B) exploration of the unique or specialized features of the Paramecium cell cycle, most of which have been revealed through analysis of the cc1 mutation [reviews 6, 14].

cycle.wmf (9604 bytes) Figure 1,

Our earlier physiological and genetic work showed the occurrence of a single major control point (PCD) late in the cell cycle (Fig. 1). At this point cells become committed to division in the present cell cycle [11, 16] and the nature of the next cell cycle (meiotic or vegetative) is set [15, 20], the duration of the G1 interval is established [17]. This differs from the typical situation in most eukaryotes, for example fission yeast [21] in which there are separate control points that gate cells into the replication at the end of the G1 period, and a separate checkpoint late in the cell cycle that gates cells into division [21]. In seeking to examine the physiological machinery underlying these control events we analyzed cdk and cyclins, common components of cell regulation in all eukaryotes and have found that there are two different classes of CDK kinases (CDK1, and CDK2) with activity peaks at IDS and PCD, respectively. The horizontal black bar in Fig. 1 indicates the macronuclear S period.

Topic 1. Molecular studies on cdks and cyclins in Paramecium

Progress . We have identified two small families of cdks and obtained full sequences of representatives of both. Similarly two classes of cyclins have been identified, full length sequences of one have been obtained.

A. Cyclin Dependent Kinases This work is the first account of CDKs in ciliates.
1. Cloning and characterization of cdk1 (Work of L. Tang in collaboration with Dr. S.L. Pelech) [publications 1,2 and 3 in which this gene is referred to as cdc2PtA]. Paramecium has two peptides (p35 and p36) that react with antibodies against the conserved PSTAIRE region of yeast CDKs. The smaller less abundant (p35) peptide binds to S. pombe p13suc1 protein and can be affinity purified. This protein has an activity pattern with a strong peak that coincides with the point of commitment to cell division. It remains active during cell division, and then drops [1]. The larger peptide does not bind p13suc1 protein, is localized largely to macronuclei and shows a pattern of activity that is highest at IDS and trails off as DNA synthesis continues. The activity of the protein parallels the pattern of macronuclear DNA synthesis, and the pattern of S-phase associated cdk activity in fission yeast [21].

cycle3.wmf (7400 bytes)  Figure 2

The p36 gene (cdk1) has been cloned [2] and encodes a protein of 308 amino acids (vs. 297 for S. pombe) and would produce a protein of ~36 KDa. Southern analysis shows that this gene is one of a family of 3 genes. A large fragment (83%) of a second cdk1 ORF was obtained that shows 96% amino acid identity with the cdk1 gene.

The most interesting and peculiar result is that both CDK classes seem to have significant in vitro enzymatic activity as monomers. In the case of the p35 peptide, gel filtration studies show that most of the activity is associated with the monomer molecular size class [1]. The larger CDK1 (p36) molecule also shows activity in the monomer size class as determined by glycerol gradient centrifugation.[22]. Immunoprecipitation experiments using antibodies to the GST-fusion protein showed no detectable co-precipitating peptides. In both of these experiments there was some enzymatic activity present in a molecular size class that would correspond to a complex of CDK with a partner molecule of about 35 kDa [1]

2. Cloning and characterization of cdk2 (Work of H. Zhang, Ph. D. Student). A complete cDNA and genomic sequence of a second class of cdc2-like gene, cdk2, has been obtained using PCR approaches starting with degenerate primers based on the sequence of cdk1 [2]. The ORF consists of 301 codons (7 shorter than CDK1). It has only 48% amino acid identity with CDK1, but shows slightly greater identity to p34 cdc2-homologues from higher eukaryotes. We believe that it is the shorter p35 peptide identified in Tang's work. It contains the full canonical PSTAIRE region, unlike CDK1, which has an amino acid substitution in this region. Genomic blotting under high stringency conditions shows that there are 2 copies of cdk2 genes in Paramecium, while there are 3 cdk1s. A single transcript of ~1.2 kb is produced at a level more than an order of magnitude lower than that for cdk1 in exponentially growing cells, but is barely detectable in starved cells.

A polyclonal antibody to the C-terminal region of the CDK2 protein has been obtained and detects a peptide of the same size as the smaller peptide detected by anti-PSTAIRE antibody. It does not cross-react with CDK1 protein. Antibodies to a GST-CDK2 fusion protein are being prepared. The low abundance, the smaller size of the molecule as well as the presence of all canonical p13suc1 binding sites [32], suggest that CDK2 is the p13suc1 binding CDK that has activity associated with commitment to division [1].

B. Cyclins. This work is the first identification, cloning and initial characterization of cyclins in ciliates
Cloning and characterization of cyclins (H. Zhang, Ph.D student). Two classes of cyclins have been isolated from the Paramecium genome by PCR using degenerate primers from the two most conserved portions of the cyclin box region [
23]. These are the first cyclins recovered in any ciliate. The full length sequences of two isoforms of one class (cln1a and cln1b) have been obtained from a macronuclear library constructed in EMBL3 (courtesy of E. Meyer) and by anchor PCR, respectively (Fig. 2). Genomic Southern analysis shows that there are only two cln1 genes. Retrieval of a full-length sequence of the cln2 class is ongoing.

The largest ORF of the cln1 sequences is 324 codons, which predicts a very small mitotic cyclin of about 37 kDa rather than the 50-60 kDa size typical of mitotic cyclins in higher eukaryotes [24]. However, in the deduced amino acid sequence the essential elements, including the 'cyclin box' (essential for interaction with the cdk partner), and the 'destruction box' (characteristic of mitotic cyclins [23] and which mediates proteolysis of cyclin by the ubiquitination pathway) can be identified. These Paramecium cyclins show homology to the cyclin box region of other cyclins (42-49% identity). However, the degree of sequence conservation is too low to unambiguously assign them to the cyclin A or B families if conservative amino acid substitutions are not considered. The short cln1 cyclin lacks the cytoplasmic localization sequence that normally lies between the destruction box and the cyclin box regions of the molecule [26]. There is a single ~1.3 kb transcript present in exponentially growing cells, but not in starved cells.

Proposal. This research focuses on functional aspects of the cloned cdk and cyclin sequences.

A. Identification of the rest of the players. 1. We will obtain full sequences for at least one cln2 gene. A number of anchored PCR and library screening approaches will be used. This will give us representatives of both classes of cyclins as well as both classes of cdks. We will also continue looking for further cyclin sequences by PCR.

2. Antibodies will be raised to both types of cyclins using either or both of synthetic peptide antigens or GST-fusion proteins, as there are sufficiently large segments of the Paramecium sequence that are free of UAA or UAG (stop) codons that are used for glutamine in Paramecium [27]

3. The availability of antibodies specific for CDK2 will make it possible to verify that CDK2 binds p13suc1 protein, and to confirm the observed kinase activity pattern, tying this gene into Tang's earlier kinase assay and gel filtration work using p13suc1 affinity purified material [1] [Fig 1].

4 Antibodies to cyclins may make it possible to determine whether additional related (cross-reacting) peptides can be identified by Western blotting in cell extracts. These will be matched against the gene squences obtained by PCR.

B.Analysis of function 1. Assay CDK2 protein levels and enzyme activity as a function of cell cycle stage by immunoprecipitating the kinase complex (cdk, cyclin and associated proteins) from cell extracts obtained from elutriated samples as done earlier with CDK1 [3].

2. In the same way protein and transcript levels for each class of cyclins will be followed through the cell cycle. In elutriation-synchronized samples. Thia is important as a benchmark for further analytical work.

3. If the antibodies to either or both the GST-CDK2 fusion protein and a GST-CLN1 fusion protein are suitable for immunoprecipitation studies, then co-IP studies will be done to determine the interacting cyclin-cdk pairs. If even one of a pair of cyclin and cdk antibodies is capable of immunoprecipitatiion, the other antibody can still be used to identify the peptides in the sedimented material. This should establish if there is a cyclin associated with enzymatically active material. It leaves open the proof of whether monomeric cdks have in vitro enzyme activity.

4. The question of whether Paramecium CDKs have activity as monomers can be approached by editing internal UAA and UAG , expression of the full length sequence as GST fusion protein in bacteria. Kinase assay will then be carried out for the GST fusion protein alone and for the protein after incubating with cell extract.. Endogenous cdk activity can be eliminated by pulling down the GST fusion CDK using glutathione agarose beads.

Topic 2. Molecular and genetic analysis of the cc1 mutation

Introduction and Progress. 1) Phenotype. Our earlier work shows that the cc1 mutant has a unique phenotype whereby it blocks commitment to division and arrests the cell cycle progression at all points prior to PCD. It also arrests macronuclear DNA replication during the S period [19]. It affects only vegetative macronuclei and not micronuclei [8] either in vegetative or sexual stages. It acts on macronuclear anlagen beginning with the earliest stages of macronuclear development [8]. We have analyzed the role of the cc1 mutant in regulation of morphogenetic processes associated with cell division in detail [1, 6, 8].

2) Kinase defect. This mutant is also associated with a substantial kinase defect; the total histone H1 kinase activity in asynchronous cc1 cell extracts is reduced by half when assayed at restrictive temperature in vitro. The cc1 mutation, however, has no significant effect on CDK1 activity (L.Tang, unpub.). This does not mean that CC1 activity is not required for normal in vivo activation.. This important observation suggests that the CC1 product is associated with the active kinase, either as the kinase itself, or more likely, as a regulatory molecule associated with it as the rescuing sequence contains no consensus kinase domain.

3) Suppressors. We have isolated several second-site suppressor mutations of cc1, one of which also suppresses another cell cycle mutant, cc3. The strong cell cycle phenotypes of the cc1 mutant, its associated kinase defect and the occurrence of interacting genetic factors all suggest that this protein is a key regulator of the cell cycle, and that it interacts with other proteins.

4) Rescue. In the last year S.M. Adl from this laboratory, in collaboration with Dr. Janine Beisson and Dr. Jean Cohen (CNRS Gif-sur-Yvette, France) has isolated by functional complementation a 1793 bp DNA fragment that strongly rescues the cc1 defect at restrictive temperature. This fragment was obtained by microinjection of a macronuclear DNA library into the macronucleus of cc1 cells that were subsequently selected for growth under restrictive conditions. After successive rounds of fractionation of the library, injection and screening selection, a single cloned sequence with rescuing activity was obtained. The largest ORF in the fragment is 281 codons. The putative amino acid sequence shows no significant similarity to anything in Genbank. The corresponding sequence has been obtained from the cc1 mutant by PCR and is being sequenced. The rescuing sequence could be a) the cc1 gene, 2) a CC1 target, or 3) other second-site suppressor locus (activator or regulator of CC1 activity). Initial PCR experiments indicate that a fragment of the same size is obtained from two other ciliates, Tetrahymena and Sterkiella, suggesting that this sequence may be conserved among ciliates. Only two other Paramecium genes have been cloned by microinjection of random DNA fragments to achieve functional complementation [28, Haynes et al, unpubl, Skouri & Cohen, unpubl].

Proposal. This molecular sequence provides a further window on the molecular control of the cell cycle in Paramecium (Proposed work of new Ph.D. student).

A. Analysis of the cc1 rescuing sequence
1) Sequencing of the corresponding sequence from cc1 cells will tell whether the mutant contains an altered DNA sequence in this region. If so, then the rescuing sequence is part of the cc1 gene, if not, it is from a regulator or co factor closely associated with the activity of cc1 (call it geneX). If the geneX sequence does not appear to be the cc1 gene, we propose to test the individual exons for rescuing activity to try to determine something more of the rescue mechanism.

2) Basic characterization In either case,(geneX is cc1, or not) we propose to a) obtain both cDNA and genomic sequences for the entire geneX; b) raise polyclonal antibodies to the GENEX protein product; c) examine the pattern of expression of geneX as a function of cell cycle stage and nutrient level; and d) examine the intracellular localization of GENEX as a function of cell cycle stage and nutrient level.

3) Analyse associated molecules and enzyme activity. a) Determine whether antibodies to GENEX can co-IP H1 kinase activity; b) determine whether any of the cdks or cyclins identified above are co-precipitated by antibodies to GENEX. c) Similarly, antibodies to the cdks and cyclins will be tested to see if they can co-IP GENEX; d) determine whether depletion of wild type cell extracts with GENEX antibodies results in a reduction in H1 kinase activity. This would be done in conjunction with 3a) above and would provide information about the intimacy of association between cc1 and the kinase defect. These experiments will be done whether geneX is cc1 or not. The interpretation of the results, will differ. The information about molecular associations is very important in either case.

B. The kinase defect
1) We will determine the cell cycle dependence of the cc1-associated H1 kinase defect by assaying kinase activity of cc1 and +/+ cells at different times in the cell cycle.

2) Determine if there is a CDK2 activity defect in cc1 cells. CDK1 activity is normal in cc1 cells, as noted above.

3) Determine whether cc1 LOF prevents normal activation of CDK1 and CDK2 histone H1 kinase activity using a temperature shift regimen in mass-synchronized cells to test whether activation of CDKs with normal timing when cc1 activity is blocked. The logic of the experiment is similar to that employed with analysis of the role of cc1 in controlling commitment to the sexual pathway [15]. This experiment will test our present assumption that cc1 acts upstream of the CDKs. The observation that CDK1 activity is unaltered in cc1 cells in vitro, mentioned above, does not preclude the possibility that CC1 activity is required upstream of CDK1 activation in vivo.

Topic 3. Cell cycle analysis in other ciliates.

A. Analysis of cell cycle regulation in Sterkiella.
( Work of Sina Adl, Ph.D. student. Publications,
4, 12 and 13). This organism has been extensively used for studies of DNA rearrangements during the development of the macronucleus [31] and called Oxytricha. Little was known about its cell cycle, or about how the vegetative cycle was adapted. Adl has developed procedures for manipulating life history events so that synchronous encystment, excystment, conjugation and morphogenetic rescaling are now possible. He has also worked out the kinetics of readjustment of size and oral replacement in response to varying nutrient conditions and prey switching. This work provides background for analysis of the molecular biology of these processes.

B. Tetrahymena. Preliminary results indicate that Tetrahymena also has two polypeptides that cross react with anti-PSTAIRE antibodies (p35 and p37). Unlike Paramecium the larger p37 molecule is the minor p13suc1 binding form and the predominant form is the smaller p35 molecule. Our immunoblotting results agree with the recently published report. [29] We have already obtained partial sequences for both a cyclin (Fig. 3), and a cdk. This project is significant because of extensive work done by Zeuthen and his colleagues on the 'division protein' of Tetrahymena e.g. [25, 30] and the failures of others in recent years to obtain cyclin-like sequences using PCR procedures. (N. Williams, pers. comm).

Proposal -CDKs and cyclins in Tetrahymena and Sterkiella. (proposed work of new M.Sc. student).This work will be initiated as part of a comparative study to determine whether the peculiarities of the Paramecium cdks are general features of ciliate cdks. We propose to initiate this project using the approaches developed for isolation and characterization of cdks and cyclins in Paramecium. Preliminary results indicate tht there are several cdks in Sterkiella and 2 in Tetrahymena.


[1]. Tang, L., S.M. Adl and J.D. Berger. A CDC2-Related Kinase is Associated with Macronuclear DNA Synthesis in Paramecium tetraurelia. J. Euk. Microbiol. 269-275. (1997),

[2] Tang, L., S.L. Pelech and J.D. Berger. Isolation of the Cell Cycle Control Gene cdc2 from Paramecium tetraurelia. Biochim. Biophys Acta. 1265, 161-167. (1995).

[3]Tang, L., S.L. Pelech and J.D Berger. A cdc2-like kinase associated with commitment to division in Paramecium tetraurelia. J. Eukaryotic Microbiology. 41, 381-386.(1994)

[4] Adl, S.M. and J.D. Berger. Timing of Life Cycle Morphogenesis in Synchronous Samples of Sterkiella histriomuscorum. I. The vegetative cell cycle. Europ. J. Protistol. 33, 99-109. (1997)

[5] Adl, S.M. and J.D. Berger. Comparison of methods of studying cell cycle progression in Paramecium tetraurelia. Journal of Eukaryotic Microbiology.42, 213-218. (1995)

[6] Adl., S.M. and J.D. Berger. Commitment to Division in Ciliate Cell Cycles. Journal of Eukaryotic Microbiology 43, 77-86 (1996).

[7] Montagnes, D.J.S. ; Berger, J.D. ; Taylor, F.J.R. Growth rate of the marine planktonic ciliate Strombidinopsis cheshiri Snyder and Ohman as a function of food concentration and interclonal variability J. EXP. MAR. BIOL. ECOL. vol. 206, 121-132 1996

[8] Adl, S.M. and J.D. Berger Cell Cycle Mutation in Paramecium tetraurelia discriminates between sexual and vegetative functions. Developmental Genetics 15, 172-175 (1994)

[9] Adl, S. M. and J.D. Berger. Timing of micronuclear mitosis and its relation to commitment to Division in Paramecium tetraurelia. Developmental Genetics 13, 234. (1992).

[10] Berger, J.D. and M. Havelka. Maternal effect mutations affecting macronuclear determination and development in Paramecium tetraurelia. Developmental Genetics 12, 226-237. (1991).

[11] Adl, S.M. and J.D. Berger. Timing or oral morphogenesis and its relation to commitment to division in Paramecium tetraurelia. Exp. Cell Res. 192, 497-504. (1991).

[12] Adl, M.S. and J.D. Berger 1997. Timing of life cycle morphogenesis in synchronous samples of Sterkiella histriomuscorum. II The sexual pathway. Archiv. Protistol. Submitted Aug 1977

[13]. Adl, S.M. and J.D. Berger. The effect of rescaling cell-size on predator-prey dynamics of sterkiella histriomuscorum in a materially closed ecosystem. Submitted October 1997..

[14] Berger (1988) The cell cycle and regulation of cell mass in Paramecium. In Görtz ed. Springer pp. 197-119.

[15] Berger (1986). Autogamy in Paramecium Exp. Cell Res. 166, 475-85.

[16] Berger & Ching (1989) ) Commitment to division in Paramecium Exp. Cell Res. 182, 90-104.

[17]. Berger & Ching (1988) The timing of initiation of DNA synthesis is established in the preceding cell cycle as cells become committed to division Exp. Cell Res. 174, 355-366.

[18] Rasmussen, Ching & Berger 1986. Effects of increased cell mass and altered gene dosage on the timing of initiation of macronuclear DNA synthesis in Paramecium Exp. Cell Res. 165, 53-62.

[19] Rasmussen and Berger (1985). ) A gene required for cell cycle progression during the G1 portion of the cell cycle and for maintenance of DNA synthesis in Paramecium Exp. Cell Res. 155, 593-597.

[20] Berger & Rahemtullah. (1990) Regulation of the sexual pathway in Paramecium. Exp. Cell Res. 187, 126-133.

[21] Stern & Nurse (1996) A quantitative model for cdc2 control of S phase and mitosis in fission yeast. Trends in Genetics 12, 345-50

[22] Tang, L. (1995). A cdc2-like protein in Paramecium tetraurelia. Ph.D. Thesis. Univ. of British Columbia.

[23] Brown et al. (1995) The crystalline structure of cyclin A. Nature - Structure 3, 1235-1247.

[24] Pines (1995) Cyclins and cyclin-dependent kinases: a biochemical view. Biochem. Journal 305, 697-711.

[25] Williams & Mackey (1991) Is cyclin Zeuthen's 'divisin protein'.? Exp. Cell Res. 197, 137-9.

[26] Pines & Hunter (1996) EMBO J. 13, 3772-13381.

[27] Caron & Meyer (1985) Does Paramecium primaurelia use a different gentic code in its macronucleus? Nature 314, 185-8, Preer et al. (1985) Deviation from the universal code shown by a gene for surface protein 51A in Paramecium. Nature 314, 188-190.

[28] Endo et al. (1995) Injection of total genomic DNA and restoration of wild-type phenotype in trichocyst non-discharge mutant TDW of Parmecium caudatum. Jpn. J. Genet. 70, 633-42.

[29] Fujishima et al. (1992) Meiosis reinitiation-inducing factor of Tetrahymena functions upstream of M-phase promoting factor. J. Protozool 39, 683-90.

[30] Watanabe & Ikeda. (1965) Further confirmatin of 'division protein' fractions in Tetrahymena pyriformis.Exp. Cell Res. 39, 464-69.

[31] Prescott (1994) DNA in hypotrich ciliates. Microbiol. Rev. 58, 233-67.

[32] Ducomun, et al. (1991) Mutations at sites involved in Suc1 binding inactivate cdc2. Mol. & Cell Biol. 11, 6177-6184.