The differential expression of the G surface antigen alleles in Paramecium primaurelia heterozygous cells correlates to macronuclear DNA rearrangement.

The Paramecium primaurelia cell surface is covered with a high molecular weight protein called the surface antigen. Several genes encode alternative surface antigens, but only one is expressed at a time. In addition, each of these genes shows a high degree of allelic polymorphism. Paramecium primaurelia strains 156 and 168 have different alleles of the G antigen gene whose respective antigens can be distinguished in vivo using specific antibodies. An interallelic exclusion phenomenon has been previously described: 94% of the 156/168 heterozygotes express only the 156 allele of the G gene; 6% express both the 156 and the 168 alleles. The phenotype of the heterozygotes is determined at the time of macronuclear differentiation. We have investigated the molecular basis for the different heterozygous phenotypes. Both mRNAs are always produced, and the 156 mRNA is always more abundant than the 168 mRNA. The relative amounts of these messages, however, vary greatly between different heterozygotes and parallel their phenotype. Pushing the analysis further, we show that the copy number of each allele in the macronucleus correlates with the relative amounts of the mRNAs. However, allelic dosage alone is not sufficient to explain the variations of the mRNA ratio. The G antigen gene is located near a telomere in the macronucleus. We show that the distance between the 156G gene and the telomere is different in homozygotes and heterozygotes. It also varies among heterozygotes and is correlated with the mRNA ratio. Thus, we have identified two different parameters, both linked to the genome rearrangements occurring during macronuclear differentiation, that correlate with the relative expression of the two alleles. Two hypotheses concerning the influence of the telomere position on the expression of the gene are discussed.     

Regulation of surface antigen expression in Paramecium primaurelia. II. Role of the surface antigen itself.

In the wild-type strains, 156 and 168, of Paramecium primaurelia, the alleles G156 and G168 expressed at medium temperature specify two immunologically distinguishable surface antigens 156G and 168G, whose phenotypic expression shows allelic exclusion, the majority of heterozygotes being phenotypically [156G] while a small minority is phenotypically [156G-168G]. At high temperature, the antigens coded by another locus, generally the D locus, are expressed. This system, displaying both intergenic and interallelic exclusion, provides favourable material to analyze the respective roles of the genome, of the antigens expressed and of the environmental conditions, in particular temperature, on the regulation of the expression of surface antigens. This analysis was carried out by studying the variations of the expression of surface antigens as a function of temperature, culture medium and previously expressed antigens in different genetic situations (a) in homozygotes: the wild-type strains 156 and 168, and the isogenized strains "G156 isogenic 168 carrying the G156 allele in a 168 genetic background; (b) in heterozygotes of the two phenotypic classes of heterozygotes, [156G] and [156G-168G]. The results show that (1) the thermal stability of the expression of a given surface antigen and its rate of re-appearance at the cell surface depend on its own specificity; (2) in heterozygotes [156G-168G], the stability of the expression of the antigen 156G is modified and "adjusted" to that of the less stable surface antigen 168G, and (3) the surface antigen itself exerts a positive control on the maintenance of its own expression. An interpretative model of "transmembranous control" is proposed to account for the regulation of the expression of surface antigens in Paramecium.     

Regulation of surface antigens expression in Paramecium primaurelia: genetic and physiological factors involved in allelic exclusion.

Summary In Paramecium aurelia, allelic exclusion can be considered as a basic feature of the surface antigens system in the same way as intergenic exclusion. Our studies on allelic exclusion in G156/G168 heterozygotes show that (1) allelic exclusion does not depend on discrete regulatory genes dispersed throughout the genome; (2) it does not seem to be influenced by cytoplasmic factors; (3) it occurs regardless of the surface antigen expressed by the parental strains at the time of the cross.These results are discussed in relation to both intergenic and interallelic exclusion for which a common basis is proposed.     

Intergenic and interallelic exclusion inParamecium primaurelia: Immunological comparisons between allelic and non-allelic surface antigens

InParamecium, the expression of surface antigens is regulated in such a way that only one is generally present at the cell surface under given environmental conditions. Previous analyses have indicated that the surface antigen molecules play a key role in the control of their own expression. In order to characterize the structural particularities displayed by both allelic and non-allelic surface antigen molecules, immunological; comparisons were performed in vivo and in immunodiffusion on nine G and six D allelic surface antigens inParamecium primaurelia.Our results show: (1) it is possible to distinguish two regions in the surface antigen molecule; one accessible to antibodies in vivo, carrying specific immobilization determinants, the other not accessible to antibodies in vivo, carrying common determinants shared by all the antigens of the same allelic series. Antigens coded by different loci differ in both regions. (2) The specificity of immobilization determinants is not borne by a hypothetical carbohydrate component of the molecule but by the polypeptide chain itself. (3) In heterozygotes displaying allelic exclusion the parental surface antigen phenotypically excluded in vivo at the cell surface is not present in the cytoplasm. These data permit some interpretations concerning the mechanisms of intergenic and interallelic exclusion, on the basis of the structural differences between the different surface antigens.I thank Doctors M. Weiss and L. Sperling very much for improvements to the English.     

Conserved sequences flank variable tandem repeats in two alleles of the G surface protein of Paramecium primaurelia

We describe the cloning and the sequencing of a macronuclear DNA fragment of Paramecium primaurelia, strain 168, encompassing the entire coding region of the 168G surface protein gene. Comparison of its nucleotide and its deduced amino acid sequences to those of the allelic surface protein 156G, previously described, reveals the rigorous conservation of a highly periodic structure. This structure is based on the presence of 37 periods of about 75 residues, each period containing eight cysteine residues. The differences between the two proteins are clustered in the central part of the sequence, which is itself made of quasi-identical tandem repeats. We propose that these repeats constitute the domain exposed on the surface of the cells and present the characteristics of concerted evolution.     

Molecular Biology of the Genes for Immobilization Antigens in Paramecium1

Several genes for surface antigens of the Paramecium aurelia complex of species have been isolated. In addition lo known deletions of the 51A gene, we have obtained deletions involving the 51B gene and have developed a procedure for obtaining deletions of additional genes. Both Mendelian and non-Mendelian deletions of both the A and B genes have been found. In the non-Mendelian deletions the genes are present in the micronuclei and absent in the macronuclei. Processing of micronuclear DNA into new macronuclear DNA at conjugation and autogamy is under the control of the old macronucleus, and newly forming macronuclei become exactly like the old. Thus in the non-Mendelian mutants, macronuclei have a specific antigen gene deleted and also are impaired in their ability to direct normal DNA processing at the next conjugation or autogamy. These cases, along with others, show that this system of macronuclear control is a fundamental feature of ciliate genetics. The sequence of the 51A and 51C genes is described and compared with the 156G and 51H genes obtained by others. The 51A and 156G genes are remarkably similar while 51Cand 51H are rather different. No introns or pseudogenes have been observed. Some, possibly all, of the genes are on the ends of chromosomes. Characteristic upstream and downstream sequences adjacent to the coding portions of the genes are given. The sequences UAA and UAG are preferred over CAA and CAG for glutamine while UGA is the true stop codon.     

Macronuclear DNA organization and transcription in Paramecium primaurelia

Macronuclear DNA was isolated from Paramecium primaurelia, stock 168. Although the macronucleus is polyploid to the extent of 840C, in other respect the DNA appears to be simply organized, having neither satellite sequences nor substantial amounts of intermediately repetitive sequence. The sequence complexity of macronuclear DNA is quite low for a eukaryote cell, being approximately 19 times more complex than the genome of Escherichia coli. In addition, the GC content is low (25%) and the isolated DNA molecules have lengths mostly in the range 0.2–5 μm. In these various respects, the macronuclear DNA of Paramecium is similar to that of other ciliates. A clone of Paramecium cultured under controlled conditions contains polyadenylated RNA sequences which are homologous to 5–8% of the macronuclear DNA. Sequence complexity analysis indicates that the polyadenylated RNA contains two abundance classes of molecules, one present at low frequency and transcribed from approximately 10⁴ genes, the other at 100 times greater concentration and transcribed from about 100 genes. The relevance of these results to the control of gene expression in Paramecium is discussed.     

Mendelian and non-mendelian mutations affecting surface antigen expression in Paramecium tetraurelia.

A screening procedure was devised for the isolation of X-ray-induced mutations affecting the expression of the A immobilization antigen (i-antigen) in Paramecium tetraurelia. Two of the mutations isolated by this procedure proved to be in modifier genes. The two genes are unlinked to each other and unlinked to the structural A i-antigen gene. These are the first modifier genes identified in a Paramecium sp. that affect surface antigen expression. Another mutation was found to be a deletion of sequences just downstream from the A i-antigen gene. In cells carrying this mutation, the A i-antigen gene lies in close proximity to the end of a macronuclear chromosome. The expression of the A i-antigen is not affected in these cells, demonstrating that downstream sequences are not important for the regulation and expression of the A i-antigen gene. A stable cell line was also recovered which shows non-Mendelian inheritance of a macronuclear deletion of the A i-antigen gene. This mutant does not contain the gene in its macronucleus, but contains a complete copy of the gene in its micronucleus. In the cytoplasm of wild-type animals, the micronuclear gene is included in the developing macronucleus; in the cytoplasm of the mutant, the incorporation of the A i-antigen gene into the macronucleus is inhibited. This is the first evidence that a mechanism is available in ciliates to control the expression of a gene by regulating its incorporation into developing macronuclei.     

Macronuclear structure of the G surface antigen gene of Paramecium primaurelia and direct expression of its repeated epitopes in Escherichia coli.

The gene encoding the G surface antigen of Paramecium primaurelia was cloned from a macronuclear DNA library by a screening procedure involving differential hybridization with cDNA probes synthesized from polyadenylated RNAs of cells expressing one of two alternate antigens. S1 mapping experiments and sequencing of the cloned DNA and the mRNA showed that the cloned gene corresponded to the high-molecular-weight mRNA that had been indirectly identified as that of the G surface antigen. Because the genetic code of Paramecium spp. is different from the "universal" code, this mRNA cannot be correctly translated in vitro; direct proof that it encoded the antigenic determinants of this protein was therefore obtained through expression of fragments of the coding sequence in Escherichia coli by using the expression vector lambda gt11. Studies on the structure of this gene revealed that the central part of the coding sequence contained at least five tandem repeats of 222 base pairs, encoding immunogenic domains of the protein. We also showed that, like other surface antigen genes of trypanosomes and paramecia, this gene lay next to a chromosome end and that no rearrangement of its immediate genomic environment was associated with its expression.     

Mutants Affecting Processing of DNA in Macronuclear Development in Paramecium

In Paramecium tetraurelia, stock 51, the A surface protein is coded by the wild type A51 gene, present in micronuclei in two copies and in macronuclei in about 1500 copies. DNA processing, comprised of DNA cleavage, copy number amplification and telomere addition occurs at autogamy and conjugation when old macronuclei degrade and new macronuclei are formed from micronuclei. In this paper we characterize mutants with macronuclear A gene deletions. These mutants are notable in three respects. First, the mutants do not appear to be simple micronuclear deletions. Although genetic analysis shows that the d12 mutant d12(-1300) is homozygous for the allele A-1300 and the mutant d12(+1) for A+1, analysis by the polymerase chain reaction indicates that the micronuclei in these two mutants contain intact, but presumably altered, micronuclear A genes. They undergo deletion during DNA processing when new macronuclei are formed. Second, the position of the deletions in these alleles has been shown to change. The deficiency present in the d12 allele A-1300 was originally determined to extend from position -1300 (relative to the start of translation of the A gene) to the end of the chromosome. Later, a derivative of this strain, homozygous for the d12 allele A+1 was isolated in which the start site of the deletion was found to have moved from -1300 to +1. Third, a surprising interaction occurs in crosses between a line homozygous for the d12 allele and one homozygous for the wild-type A51 allele. Previous work on the non-Mendelian d48 mutant (which has intact A51 genes in its micronucleus, but has truncated A51 genes in its macronucleus) has shown that intact A51 alleles must be present in the old macronucleus in order for A51 alleles to undergo proper processing. We find that d12 alleles act on A51 alleles in heterozygotes such that intact macronuclear A genes are no longer required for proper processing of A51. Thus, in crosses of 51 x d12 (either +1 or -1300) d12 exconjugants, as well as 51 exconjugants, give rise to clones carrying both intact A51 and truncated d12 alleles. Remarkably the d12 alleles, which are themselves deleted during processing, are capable in the heterozygote of fostering normal processing of the A51 allele.     

Developmentally regulated chromosome fragmentation linked to imprecise elimination of repeated sequences in paramecia

The chromosomes of ciliates are fragmented at reproducible sites during the development of the polyploid somatic macronucleus, but the mechanisms involved appear to be quite diverse in different species. In Paramecium aurelia, the process is imprecise and results in de novo telomere addition at locally heterogeneous positions. To search for possible determinants of chromosome fragmentation, we have studied an ~21-kb fragmentation region from the germ line genome of P. primaurelia. The mapping and sequencing of alternative macronuclear versions of the region show that two distinct multicopy elements, a minisatellite and a degenerate transposon copy, are eliminated by an imprecise mechanism leading either to chromosome fragmentation and the formation of new telomeres or to the rejoining of flanking sequences. Heterogeneous internal deletions occur between short direct repeats containing TA dinucleotides. The complex rearrangement patterns produced vary slightly among genetically identical cell lines, show non-Mendelian inheritance during sexual reproduction, and can be experimentally modified by transformation of the maternal macronucleus with homologous sequences. These results suggest that chromosome fragmentation in Paramecium is the consequence of imprecise DNA elimination events that are distinct from the precise excision of single-copy internal eliminated sequences and that target multicopy germ line sequences by homology-dependent epigenetic mechanisms.     

Developmentally Controlled Rearrangement of Surface Protein Genes in Paramecium tetraurelia1

ABSTRACT Early research on Paramecium genetics highlighted the role of the cytoplasm on inheritance. Today this tradition continues as recent investigations of macronuclear development in Paramecium have revealed unusual cytoplasmic effects that are not easily explained within current paradigms. It is generally assumed that most programmed DNA rearrangements in ciliates are regulated by cis acting signals encoded within the germline (micronuclear) DNA, but there are increasing examples in which the old macronucleus acts through the cytoplasm (in trans) to affect the loss and rearrangement of DNA in the developing macronucleus. The remarkable specificity of this effect has forced a reevaluation of the standard view of macronuclear determination in Paramecium. This review summarizes our knowledge of the effect of the old macronucleus on the developmentally controlled rearrangements of the P. tetraurelia, stock 51A and B variable surface protein genes.     

Sequence-specific epigenetic effects of the maternal somatic genome on developmental rearrangements of the zygotic genome in Paramecium primaurelia.

In ciliates, the germ line genome is extensively rearranged during the development of the somatic macronucleus from a mitotic product of the zygotic nucleus. Germ line chromosomes are fragmented in specific regions, and a large number of internal sequence elements are eliminated. It was previously shown that transformation of the vegetative macronucleus of Paramecium primaurelia with a plasmid containing a subtelomeric surface antigen gene can affect the processing of the homologous germ line genomic region during development of a new macronucleus in sexual progeny of transformed clones. The gene and telomere-proximal flanking sequences are deleted from the new macronuclear genome, although the germ line genome remains wild type. Here we show that plasmids containing nonoverlapping segments of the same genomic region are able to induce similar terminal deletions; the locations of deletion end points depend on the particular sequence used. Transformation of the maternal macronucleus with a sequence internal to a macronuclear chromosome also causes the occurrence of internal deletions between short direct repeats composed of alternating thymines and adenines. The epigenetic influence of maternal macronuclear sequences on developmental rearrangements of the zygotic genome thus appears to be both sequence specific and general, suggesting that this trans-nucleus effect is mediated by pairing of homologous sequences.     

Delayed degradation of parental macronuclear DNA in programmed nuclear death of Paramecium caudatum

In the ciliated protozoan Paramecium caudatum, a parental macronucleus that is fragmented into some 40-50 pieces during conjugation does not degenerate immediately, but persists until the eighth cell cycle after conjugation. Here we demonstrate that the initiation of the parental macronuclear degeneration occurs at about the fifth cell cycle. The size of parental macronuclear fragments continued to increase between the first and fourth cell cycle, but gradually decreased thereafter. By contrast, a new macronucleus grew and reached a maximum size by the fourth cell cycle, suggesting that the new macronucleus matured by that stage. Southern blot analysis revealed that parental macronuclear DNA was degraded at about the fifth cell cycle. The degradation was supported by acridine orange staining, indicating degeneration of the macronuclear fragments. Prior to the degradation, the fragments once attached to the new macronucleus were subsequently liberated from it. These observations lead us to conclude that once a new macronucleus has been fully formed by the fourth cell cycle, the parental macronuclear fragments are destined to degenerate, probably through direction by new macronucleus. Considering the long persistence of the parental macronucleus during the early cell cycles after conjugation, the macronuclear fragments might function in the maturation of the imperfect new macronucleus. Two possible functions, a gene dosage compensation and adjustment of ploidy level, are discussed.     

Identification of the Immobilization Antigens of Paramecium by Their Cyanogen Bromide Cleavage Patterns1

Immobilization antigens from 12 serotypes of three stocks of Paramecium tetraurelia and from one serotype of one stock of P. primaurelia were isolated and purified. Purified proteins were cleaved with cyanogen bromide, and the patterns of the fragment peptides were determined by electrophoresis on SDS-polyacrylamide gels. It was shown that each of the serotypes of stock 51 of P. tetraurelia has an antigen that produces a characteristic and unique pattern. Consequently, the antigens can be identified by their patterns. Antigens from the allelic serotypes tested had identical patterns. The method is sensitive enough for the investigation of small sample volumes, and useful as a simple biochemical technique for the identification of serotypes.     

Molecular biology of the genes for immobilization antigens in Paramecium

Several genes for surface antigens of the Paramecium aurelia complex of species have been isolated. In addition to known deletions of the 51A gene, we have obtained deletions involving the 51B gene and have developed a procedure for obtaining deletions of additional genes. Both Mendelian and non-Mendelian deletions of both the A and B genes have been found. In the non-Mendelian deletions the genes are present in the micronuclei and absent in the macronuclei. Processing of micronuclear DNA into new macronuclear DNA at conjugation and autogamy is under the control of the old macronucleus, and newly forming macronuclei become exactly like the old. Thus in the non-Mendelian mutants, macronuclei have a specific antigen gene deleted and also are impaired in their ability to direct normal DNA processing at the next conjugation or autogamy. These cases, along with others, show that this system of macronuclear control is a fundamental feature of ciliate genetics. The sequence of the 51A and 51C genes is described and compared with the 156G and 51H genes obtained by others. The 51A and 156G genes are remarkably similar while 51C and 51H are rather different. No introns or pseudogenes have been observed. Some, possibly all, of the genes are on the ends of chromosomes. Characteristic upstream and downstream sequences adjacent to the coding portions of the genes are given. The sequences UAA and UAG are preferred over CAA and CAG for glutamine while UGA is the true stop codon.     

Telomeric site position heterogeneity in macronuclear DNA of Paramecium primaurelia.

In Paramecium primaurelia, the macronuclear gene encoding the G surface protein is located near a telomere. In this study, multiple copies of this telomere have been isolated and the subtelomeric and telomeric regions of some of them have been sequenced. The telomeric sequences consist of tandem repeats of the hexanucleotides C4A2 or C3A3. We show that the location where these repeats are added, which we call the telomeric site, is variable within a 0.6-0.8-kb region. These results are discussed in relation with the formation of macronuclear DNA.     

Analysis of mating type determination by transplantation of O macronuclear karyoplasm in Paramecium tetraurelia

The odd (O) or even (E) mating type in Paramecium tetraurelia is determined during the first cell cycle after new macronuclear development. The present paper demonstrates that mating type E is irreversibly determined at the end of the first cell cycle. Direct evidence comes from transplanting O macronuclear karyoplasm containing O-determining factor into E autogamous cells during a new postzygotic macronuclear development. Transplantation of O macronuclear karyoplasm into E autogamous cells at 7–8 hr after the origin of the macronucleus from a product of the synkaryon produces nearly 100% O mating type among the exautogamous cell lines but almost none 10–11 hr after the origin of the macronucleus (around the end of the first cell cycle). The macronuclear anlagen at the stage at which mating type E seems to be fixed contains about 20 times as much DNA as the vegetative G1 micronucleus. The O-determining factor shifting E cells toward O mating type by transplanting O macronuclear karyoplasm is also produced by the newly developed macronucleus in an effective concentration at 10–11 hr after the sensitive period and produced at full levels by the third cell cycle. The level of O factor in the macronucleus then gradually declines with subsequent repeated rounds of DNA synthesis and is finally lost by the eighth cell cycle.