STUDIES ON NUCLEAR STRUCTURE AND FUNCTION IN TETRAHYMENA PYRIFORMIS : I. RNA Synthesis in Macro- and Micronuclei

Tetrahymena in the log phase of growth were pulse labeled with uridine-³H, fixed in acetic-alcohol, extracted with DNase, and embedded in Epon. 0.5-µ sections were cut, coated with Kodak NTB-2 emulsion, and developed after suitable exposures. Grains were counted above macronuclei, above 1000 micronuclei, and above 1000 micronucleus-sized "blanks" which were situated next to micronuclei in the visual field by means of a camera lucida. An analysis of grain counts showed that micronuclei were less than ½₀₀₀ as active as macronuclei on the basis of grains per nucleus. Since micronuclei contained, on the average, about ½₀ as much DNA as macronuclei, micronuclear DNA had less than 1% of the specific activity of macronuclear DNA in RNA synthesis. However, even this small amount of apparent incorporation was not significantly different from zero. Comparisons of the frequency distributions of labeled micronuclei with those of micronuclear "blanks" showed no evidence of a small population of labeled nuclei such as might be expected if micronuclei synthesized RNA for only a brief portion of the cell cycle. We conclude from these studies that there is no detectable RNA synthesis in Tetrahymena micronuclei during vegetative growth and reproduction.     

[6N]METHYL ADENINE IN THE NUCLEAR DNA OF A EUCARYOTE, TETRAHYMENA PYRIFORMIS

DNA isolated from macronuclei of the ciliate, Tetrahymena pyriformis, has been found to contain [⁶N]methyl adenine (MeAde); this represents the first clear demonstration of significant amounts of MeAde in the DNA of a eucaryote. The amounts of macronuclear MeAde differed slightly between different strains of Tetrahymena, with approximately 0.65–0.80% of the adenine bases being methylated. The MeAde content of macronuclear DNA did not seem to vary in different physiological states. The level of MeAde in DNA isolated from micronuclei, on the other hand, was quite low (at least tenfold lower than in macronuclear DNA).     

Morphology and development of the macronuclei of the ciliates Stylonychia mytilus and Euplotes aediculatus

Some stages of macronuclear anlagen development, known from earlier investigations (see Fig. 1), were studied in detail. The results are: a) The giant chromosomes of Stylonychia mytilus are not somatically paired, but are connected end-to-end to form one or a few composite chromosomes. When they later disintegrate, the bands become isolated granules. b) Spectrophotometric measurements show that during the DNA-poor stage which follows the disintegration of the chromosomes, the macronuclear anlagen of Euplotes have a DNA content of 21 c, while the syncaryotic (deriving from syncarya) and hemicaryotic (deriving from haploid hemicarya) anlagen of Stylonychia have the DNA content of diploid micronuclei (2c). Nevertheless the syncaryotic anlagen of Stylonychia and Euplotes initially develop two nucleoli at the end of this stage, the hemicaryotic anlagen of Stylonychia only one. From this it is concluded that the genes of one giant chromosome band stay together in one granule, c) Labeled DNA from the giant chromosomes which remains in the anlagen during the DNA-poor stage is distributed approximately equally to the daughter nuclei during the first few fissions of the exconjugants.-Autoradiographic experiments showed that the DNA of the macronuclei of Stylonychia that is duplicated at one time in a replication band is not duplicated simultaneously during the next DNA-duplication. The DNA duplications during the second polyploidization stage of the macronuclear anlagen development are exceptions, because the mixing of the macronuclear DNA which occurs before every fission does not occur during the second polyploidization stage.—The pseudomicronuclei which sometimes are formed from the macronuclei in emicronucleated strains of Stylonychia contain numerous elements which are much smaller than the chromosomes.—The macronucleus of Stylonychia is very insensitive to irradiation with X-rays.—The results lead to the following hypothesis: The macronuclei of the two hypotrich ciliates contain unconnected chromomeres or small aggregates which are distributed at random to the two daughter nuclei during the divisions.Research supported by the Deutsche Forschungsgemeinschaft.     

DNA amounts in the nuclei ofParamecium aurelia andTetrahymena pyriformis

DNA amounts have been determined in the micronuclei and macronuclei of 8 strains ofParamecium aurelia and 6 strains ofTetrahymena pyriformis. In the case ofTetrahymena a distribution of values for the amount of DNA in the macronuclei of all the strains was observed but the lowest values were approximately the same, viz. 1.17×10⁻¹¹ g. There are two groups of strains in relation to micronuclear DNA values ofTetrahymena, one giving an average of 0.36×10⁻¹² g and the other 0.815×10⁻¹² g. The ratio of MIC/MAC DNA varies in the two groups.Paramecium again has a range of macronuclear values within each stock—lowest value 2.51×10⁻¹⁰ g—and the micronuclear values are similar in all stocks—approximately 0.613×10⁻¹² g. The ratio of MIC/MAC DNA is similar in each stock.—The haploid genome values calculated from these data show excellent agreement with the values obtained by DNA renaturation studies. Supported by a Research Grant B/SR/8276 from the Science Research Council. The Vickers densitometer was purchased with a grant from the Medical Research Council.     

Characterization of macronuclear DNA in five species of ciliates

Macronuclear DNA of four hypotrichous and one holotrichous ciliate species was characterized by biochemical techniques. The renaturation kinetics of the macronuclear DNAs of all five species were similar. Repetitive sequences occur only in an amount below 2%. Although the DNA content of the macronuclei of the species differs considerably, the kinetic complexity of the macronuclear DNA is rather uniform (around 3 × 10¹⁰ daltons, i.e., 4–11 x the E. coli genome). Only in the macronuclei of the hypotrichous species the DNA exists as gene-sized fragments.Dedicated to Professor W. Beermann on the occasion of his 60th birthday     

Organization of the 5S RNA genes in macro- and micronuclei of Tetrahymena pyriformis

The organization of the 5S genes in macro- and micronuclei of Tetrahymena pyriformis was studied using restriction endonucleases. After complete digestion of macronuclear DNA with BamH-I or Hpa I, 5S RNA hybridized to a DNA fragment of approximately 280 base pairs (bp). When macronuclear DNA was only partially digested with these enzymes, hybridization with ³²P-5S RNA demonstrated an oligomeric series with a spacing of 280 bp. These results indicate that the 5S genes are tandemly repeated in macronuclei and that the repeating unit is 280 bp (or 180,000 daltons). Since 5S RNA is 120 nucleotides, we conclude that the 5S repeat units contain a 120 bp transcribed region and a 160 bp spacer region. When macronuclear DNA was digested with Eco RI, Bgl I, or Eco RI + Bgl I, 5S RNA hybridized to DNA of molecular weight 3–4×10⁶, suggesting that these enzymes do not cleave within a 5S repeat. These 3–4×10⁶ dalton fragments define the maximum size of an average cluster of 5S repeated units. Assuming the size of the 5S repeat to be 0.18×10⁶ daltons, there are about 15–20 5S repeats per average tandem cluster, and since there are 350 5S-genes per haploid genome, there must be approximately 15–20 tandem arrays. Results obtained using micronuclear DNA suggest that organization of the 5S-genes is very similar in macro- and micronuclei. Macronuclear rRNA genes are extracnromosomal palindromic dimers. In contrast, 5S genes in Tetrahymena were found to be integrated within the genomes of both macro- and micronuclei and not linked to the rRNA genes. Moreover, it is unlikely that they are palindromes; rather they appear to be tandemly repeated in head-to-tail linkages. Thus, the organization of the 5S genes in Tetrahymena is similar to that of higher eukaryotes.     

Characterization of the macronuclear DNA of different species ofTetrahymena

Summary The macronuclear DNAs from 20 different species ofTetrahymena were characterized using alternating Orthogonal Field (AOF) gel electrophoresis. Each species has approximately 300 different macronuclear DNA molecules that range in size from about 100–2000 kb pairs. Although the individual macronuclear DNA molecules are not well resolved on an AOF gel, most species have a unique profile of macronuclear DNA. The sequences that hybridize with histone H4 (Tetrahymena) and ubiquitin (yeast) genes were identified on the separated macronuclear DNA molecules of the different species. All species have 2 histone H4 genes located on macronuclear DNA molecules of different sizes. This is consistent with the duplication of the histone H4 gene prior to the speciation events leading to the various species ofTetrahymena. The number and sizes of the macronuclear DNA molecules that hybridize with the ubiquitin probe vary from species to species. A grouping of the different species ofTetrahymena based on this hybridization pattern paralels groupings of the species based on ribosomal RNA sequences and isoenzymes. Some intraspecific variation among different strains ofTetrahymena thermophila was detected using ubiquitin and 5S ribosomal RNA as probes.Presented at the FEBS Symposium on Genome Organization and Evolution, held in Crete, Greece, September 1–5, 1986     

Numbers of 5S and tRNA genes in macro- and micronuclei of Tetrahymena pyriformis

Macronuclei of Tetrahymena pyriformis contain approximately 200 copies of the genes for 25S and 17S ribosomal RNA (rRNA) per haploid genome. Micronuclei, however, contain only a few copies of the rRNA genes per haploid complement. Since macronuclei develop from, products of meiosis, fertilization and division of micronuclei, we suggested that the multiple copies of the rRNA genes in macronuclei are generated by amplification of the small number of genes in micronuclei (Yao et al., 1974). This process provides a simple mechanism for maintaining the homogeneity of the repeated rRNA genes. To test if amplification is a general mechanism operating on all repeated genes in Tetrahymena, we have examined the numbers of 5S RNA and tRNA genes in macro- and micronuclei. 5S RNA was purified by polyacrylamide gel electrophoresis and hybridized to saturation against macro- and micronuclear DNA. Approximately 0.013–0.014% of macronuclear DNA and about 0.009% of micronuclear DNA is complementary to 5S RNA. After correcting for the differences in the DNA sequence complexities between the two nuclei, we calculate that there are 300–350 5S genes per haploid macro- or micronuclear genome. From these data we conclude that there is little or no detectable amplification of the 5S genes in macronuclei relative to micronuclei. Similar studies using tRNA indicate that these genes are also highly repeated in both nuclei; about 800 genes are present per haploid genome. Thus, amplification from a small number of genes can be excluded as the mechanism for generating the repeated copies of the 5S and tRNA genes in Tetrahymena and it is likely that another, as yet unidentified, mechanism operates to maintain the homogeneity of these genes.     

Nuclear differentiation and nucleic acid synthesis in well-fed exconjugants of Paramecium aurelia

Paramecium aurelia exconjugants contain new macronuclear anlagen and numerous fragments of the old pre-zygotic macronucleus. Macronuclear anlagen develop during the first two cell cycles after conjugation. During this time their volume increases from about 11 m³ to about 3700 m³ and more than 10 doublings of DNA content occur. The rate of DNA synthesis is between two and three times as great as in the vegetative macronucleus. — In macronuclear fragments, however, DNA synthesis is suppressed. The rate of DNA synthesis in macronuclear fragments during the extended first cell cycle after conjugation (11 1/2 hr. vs. 5 1/2 hr. for the vegetative cell cycle) is only about one-third of the rate in vegetative macronuclei and there is only a 65% increase in the mean DNA content of fragments. The rate of fragment DNA synthesis continues to decrease during each of the subsequent two cell cycles. — Unlike the rate of DNA synthesis, the rate of RNA synthesis per unit of DNA is similar in macronuclear anlagen, macronuclear fragments and fully developed macronuclei. Macronuclear fragments continue to synthesize RNA at the normal rate long after the new macronuclei are fully developed. Fragments contribute about 80% of all RNA synthesized during the first two cell cycles after conjugation. RNA synthesis begins very early in the development of macronuclear anlagen and nucleolar material appears during the first half-hour of anlage development. — Chromosome-like structures were never observed during anlage development and there was no evidence of two periods of DNA synthesis separated by a DNA poor stage as has been observed in several hypotrichous Ciliates.     

A Nuclear Protein Involved in Apoptotic-like DNA Degradation in Stylonychia: Implications for Similar Mechanisms in Differentiating and Starved Cells

Ciliates are unicellular eukaryotic organisms containing two types of nuclei: macronuclei and micronuclei. After the sexual pathway takes place, a new macronucleus is formed from a zygote nucleus, whereas the old macronucleus is degraded and resorbed. In the course of macronuclear differentiation, polytene chromosomes are synthesized that become degraded again after some hours. Most of the DNA is eliminated, and the remaining DNA is fragmented into small DNA molecules that are amplified to a high copy number in the new macronucleus. The protein Pdd1p (programmed DNA degradation protein 1) from Tetrahymena has been shown to be present in macronuclear anlagen in the DNA degradation stage and also in the old macronuclei, which are resorbed during the formation of the new macronucleus. In this study the identification and localization of a Pdd1p homologous protein in Stylonychia (Spdd1p) is described. Spdd1p is localized in the precursor nuclei in the DNA elimination stage and in the old macronuclei during their degradation, but also in macronuclei and micronuclei of starved cells. In all of these nuclei, apoptotic-like DNA breakdown was detected. These data suggest that Spdd1p is a general factor involved in programmed DNA degradation in Stylonychia.     

STUDIES ON NUCLEAR STRUCTURE AND FUNCTION IN TETRAHYMENA PYRIFORMIS : II. Isolation of Macro- and Micronuclei

This report describes a rapid, efficient method for isolating macronuclei from Tetrahymena. The macronuclear fraction contains only small amounts of micronuclear material and little detectable whole cell or cytoplasmic contamination. A method is also described for preparing a "micronuclear fraction" which contains 20–40 micronuclei for every macronucleus present. Electron microscope observations indicate that the ultrastructure of the nuclei in the macronuclear fraction closely resembles that of nuclei in situ. The presence of ribosomes on the outer membrane of micronuclei and of pores in the micronuclear envelope is also described.     

Thermotropic ‘two-stage’ liquid crystalline crystalline lipid phase separation in microsomal membranes

The effect of temperature on native microsomal membrane vesicles isolated from Tetrahymena is investigated by wide angle X-ray diffraction. A 4.2 Å reflection, typical for lipids in the crystalline state, can be recorded in the temperature range between 0°C and 35°C. Quantitative evaluation of this reflection reveals a broad thermotropic ‘two-stage’ liquid crystallinecrystalline lipid phase separation with a ‘breakpoint’ at approx. 18°C. This ‘breakpoint’ coincides with the emergence of lipid-protein segregations in endomembranes of intact Tetrahymena cells as previously visualized by freeze-etch electron microscopy.     

Methylation of adenine in the nuclear DNA of Tetrahymena is internucleosomal and independent of histone H1

There are about 50 copies of each chromosome in the somatic macronucleus of the ciliated protozoan Tetrahymena. Approximately 0.8% of the adenine residues in the macronuclear DNA of Tetrahymena are methylated to N⁶-methyladenine. The degree of methylation varies between sites from a very low percentage to >90%. In this study a correlation was found between nucleosome positioning and DNA methylation. Eight GATC sites with different levels of methylation were examined. There was a direct correlation between the degree of methylation and proximity to linker DNA at these sites. Although methylation occurs preferentially in linker DNA, the patterns and extent of methylation in a histone H1 knockout strain were virtually indistinguishable from those in wild-type cells.     

LIA4 Encodes a Chromoshadow Domain Protein Required for Genomewide DNA Rearrangements in Tetrahymena thermophila

Extensive DNA elimination occurs as part of macronuclear differentiation during Tetrahymena sexual reproduction. The identification of sequences to excise is guided by a specialized RNA interference (RNAi) machinery that targets the methylation of histone H3 lysine 9 (K9) and K27 on chromatin associated with these internal eliminated sequences (IESs). This modified chromatin is reorganized into heterochromatic subnuclear foci, which is a hallmark of their subsequent elimination. Here, we demonstrate that Lia4, a chromoshadow domain-containing protein, is an essential component in this DNA elimination pathway. LIA4 knockout (ΔLIA4) lines fail to excise IESs from their developing somatic genome and arrest at a late stage of conjugation. Lia4 acts after RNAi-guided heterochromatin formation, as both H3K9 and H3K27 methylation are established. Nevertheless, without LIA4, these cells fail to form the heterochromatic foci associated with DNA rearrangement, and Lia4 accumulates in the foci, indicating that Lia4 plays a key role in their structure. These data indicate a critical role for Lia4 in organizing the nucleus during Tetrahymena macronuclear differentiation.     

Histone rearrangements accompany nuclear differentiation and dedifferentiation in Tetrahymena

Histone synthesis and deposition into specific classes of nuclei has been investigated in starved and conjugating Tetrahymena. During starvation and early stages of conjugation (between 0 and 5 hr after opposite mating types are mixed), micronuclei selectively lose preexisting micronuclear-specific histones α, β, γ, and H3^F. Of these histones, only α appears to accumulate in micronuclear chromatin through active synthesis and deposition during the mating process. Curiously, α is not observed (by stain or label) in young macronuclear anlagen (4C, 10 hr of conjugation). Thus, young macronuclear anlagen are missing all of the histones which are known to be specific to micronuclei of vegetative cells. By 14–16 hr of conjugation, we observe active synthesis and deposition of macronuclear-specific histones, hv1, hv2, and H1, into new macronuclear anlagen (8C). Thus macronuclear differentiation seems well underway by this time of conjugation. It is also in this time period (14–16 hr) that we first detect significant amounts of micronuclear-specific H1-like polypeptides β and γ in micronuclear extracts. These polypeptides do not seem to be synthesized during this period, which suggests that β and γ are derived from a precursor molecule(s). Since these micronuclear-specific histones do not appear in micronuclear chromatin until after other micronuclei have been selected to differentiate as macronuclei, we suspect that micronuclear differentiation is also an important process which occurs in 10–16 hr mating cells. Our results also suggest that proteolytic processing of micronuclear H3^S into H3^F (which occurs in a cell cycle dependent fashion during vegetative growth) is not operative during most if not all of conjugation. Thus micronuclei of mating cells contain only H3^S which also seems consistent with the fact that some micronuclei differentiate into new macronuclei (micronuclear H3^S is indistinguishable from macronuclear H3). Interestingly, the only H3 synthesized and deposited into the former macronucleus of mating cells is the relatively minor macronuclear-specific H3-like variant, hv2. These results demonstrate that significant histone rearrangements occur during conjugation in Tetrahymena in a manner consistent with the fact that during conjugation some micronuclei eventually differentiate into new macronuclei. Our results suggest that selective synthesis and deposition of specific histones (and histone variants) plays an important role in the nuclear differentiation process in Tetrahymena. The disappearance of specific histones also raises the possibility that developmentally regulated proteolytic processing of specific histones plays an important (and previously unsuspected) role in this system.     

Elimination of germ-line tandemly repeated sequences from the somatic genome of the ciliate Oxytricha fallax

The ciliated protozoa exhibit nuclear dimorphism. The genome of the somatic macronucleus arises from the germ-line genome of the micronucleus following conjugation. We have studied the fates of highly repetitious sequences in this process. Two cloned, tandemly repeated sequences from the micronucleus of Oxytricha fallax were used as probes in hybridizations to micronuclear and macronuclear DNA. The results of these experiments show: (1) the cloned repeats are members of two apparently unrelated repetitious sequence families, which each appear to comprise a few percent of the micronuclear genome, and (2) the amount of either family in the macronuclei from which our DNA was prepared is about 1/15 that found in an equal number of diploid micronuclei. Most, if not all, of the apparent macronuclear copies of these repeats can be accounted for by micronuclear contamination, which strongly suggests that these sequences are eliminated from the macronuclei and have no vegetiative function.     

Genome structure of Tetrahymena pyriformiis

Reassociation kinetics of DNA from the macronucleus of the ciliate, Tetrahymena pyriformis GL, has been studied. The genome size determined by the kinetic complexity of DNA was found to be 2.0×10⁸ base pairs (or 1.2×10¹¹ daltons). About 90% of the macronuclear DNA fragments 200–300 nucleotides in length reassociate at a rate corresponding to single-copy nucleotide sequences, and 7–9% at a rate corresponding to moderate repetitive sequences; 3–4% of such DNA fragments reassociate at C₀t practically equal to zero. To investigate the linear distribution of repetitive sequences, DNA fragments of high molecular weight were reassociated and reassociation products were treated with Sl-nuclease. DNA double-stranded fragments were then fractionated by size. It has been established that in the Tetrahymena genome long regions containing more than 2000 nucleotides make up about half of the DNA repetitive sequences. Another half of the DNA repetitive sequences (short DNA regions about 200–300 nucleotides long) intersperse with single-copy sequences about 1,000 nucleotides long. Thus, no more than 15% of the Tetrahymena genome is patterned on the principle of interspersing single-copy and short repetitive sequences. Most of the so called zero time binding or foldback DNA seem to be represented by inverted self-complementary (palindromic) nucleotide sequences. The conclusion has been drawn from the analysis of this fraction isolated preparatively by chromatography. About 75% of the foldback DNA is resistant to Sl-nuclease treatment. The Sl-nuclease resistance is independent of the original DNA concentration. Heat denaturation and renaturation are reversible and show both hyper and hypochromic effects. The majority of the inverted sequences are unique and about 20% are repeated tens of times. According to the equilibrium distribution in CsCl density gradients the average nucleotide content of the palindromic fraction does not differ significantly from that of total macronuclear DNA. It was shown that the largest part of this fraction of the Tetrahymena genome are not fragments of ribosomal genes.     

Selective inhibition of DNA synthesis in macronuclear fragments in Paramecium aurelia exconjugants and its reversal during macronuclear regeneration

In Paramecium exconjugants very rapid DNA synthesis takes place in the developing macronuclear anlagen, while DNA synthesis is suppressed in macronuclear fragments. The rate of DNA synthesis in fragments (as a percentage of the rate in anlagen or macronuclei in the same cells) decreases by about 40% during each successive cell cycle over at least the first five cell cycles after conjugation, even though macronuclear anlagen are fully mature by the end of the second cell cycle. — Suppression of DNA synthesis in macronuclear fragments is reversible. If macronuclear anlagen are removed at fission, a very high rate of DNA synthesis resumes in macronuclear fragments after a two-hour lag. The total rate of synthesis in the ensemble of macronuclear fragments in cells without anlagen is greater than that in anlagen in control cells. Thus, suppression of DNA synthesis in macronuclear fragments is not the result of any stable differentiation or irreversible change in the fragments but is the result of, and dependent on, the presence of macronuclear anlagen. — The results of injection of cytoplasm from vegetative cells into normal exeonjugants suggest that normal macronuclei produce an inhibitor which selectively suppresses DNA synthesis in macronuclear fragments. In control cells the relative rate of DNA synthesis in fragments ranged from 40 to 70% of that in anlagen in the same cells, while in injected cells the relative rate of incorporation of DNA precursors was suppressed to as little as 7%. The mean level of incorporation into fragments in injected cells was significantly lower than that in controls, suggesting that the injected cytoplasm contained an inhibitor.Contribution 822, Zoology Department, Indiana University. Supported in part by contract COO-235-66 of the USAEC and by grant No. Gm 15410-05 of the USPHS to T. M. Sonneborn.This paper is a portion of a dissertation submitted in partial fulfillment of the equirements for the degree of Doctor of Philosophy.     

Studies on the Macronuclei of Doublet Paramecium tetraurelia: Distribution of Macronuclei and DNA at Fission*

SYNOPSIS. Doublet Paramecium tetraurelia would be expected to contain 2 macronuclei if their nuclear complement were strictly analogous to that of singlets. However, most doublets are unimacronucleate. It is shown in this study that dimacronucleate cells are present only in young clones. Unimacronucleate cells arise either through abnormalities in the determination and distribution of macronuclear anlagen during the first cell cycle after conjugation, or from dimacronucleate cells through abnormal division and segregation of macronuclei during the fission process. When a change in the number of macronuclei occurs through abnormalities in the division and segregation of daughter macronuclei, the daughter cells produced typically have DNA contents more similar than those expected from either random segregation of daughter macronuclei, or from the normal segregation pattern in ciliates in which changes in the number of macronuclei in progeny cells do not occur. This suggests that part of the regulation process of macronuclear DNA content in Paramecium may occur through control of the segregation pattern of daughter macronuclei.     

Low molecular weight DNA in the heteromeric macronuclei of two cyrtophorid ciliates

Summary— The size range of the native DNA molecules in the heteromeric macronuclei of two cyrtophorid ciliates (Trithigmostoma cucullulus, Chilodonella uncinata) was mainly investigated by using agarose gel electrophoresis. Numerous bands superimposed on a continuous spectrum of molecular sizes between about 0.35 kb and 30 kb were resolved by conventional electrophoresis. Species-specific banding patterns indicate a variation between species in the copy number of individual DNA fragments. A slight intra-specific variability of banding patterns can exist. Electrophoretic distributions for two strains of T cucullulus were indeed found to differ by at least one more intense band (‘overamplified’ sequences?). Fractionation by contour-clamped homogeneous electric field (CHEF) gel electrophoresis revealed that the size continum of macronuclear DNA molecules does not extend beyond 60–70 kb. The average size was estimated to be around 4 kb. Unresolved DNA fraction (> 1000 kb) accounted for less than 10% of the mass of cellular DNA entering CHEF gels. Macronuclear ribosomal DNA of each cyrtophorid species was identified by Southern hybridization with a Tetrahymena rDNA probe. The hybridization signal was observed on a single band of low molecular weight DNA. The corresponding size was close to 14.5 kb in Trithigmostoma and 15.5 kb in Chilodonella, which is about twice the size of monomeric rDNA in hypotrichous ciliates. We showed that S1 nuclease resistant duplexes wit half the length of the native rDNA can be formed by rapid renaturation of heat-denatured molecules and hybridized with native rDNA. This strongly suggests that the nucleotide sequence of this rDNA is a large palindrome. Unlike the hypotrichs, macronuclear rDNA in cyrtophorids should be organized into palindromic dimers as in Tetrahymena species.