Dna from isolated pellicles of Tetrahymena

The existence of specific mitochondrial and chloroplast DNA’s is now well established but, in contrast, the presence of DNA in centrioles or basal bodies (kinetosomes) has been the subject of many conflicting reports. Studies by Randall & Disbrey (1965) showed the presence of DNA in isolated pellicles of Tetrahymena pyriformis by staining with acridine orange and by autoradiography. Smith-Sonneborn & Plaut (1967) obtained similar results with Paramecium aurelia. Chemical estimations of the amount of DNA present in isolated pellicles of Tetrahymena were made by Seaman (1960) and Hoffman (1965). Recently, Hufnagel (1969) partially characterized the DNA from isolated pellicles of Paramecium aurelia and found that it was of the same buoyant density as either the nuclear DNA or the DNA isolated from bacteria used as the food source for the cultures. No detailed physical study on the properties of DNA from pellicles of Tetrahymena has been reported. Since detailed knowledge of the physical properties of the mitochondrial and nuclear DNA’s of Tetrahymena already exists (Suyama, 1966; Flavell & Jones, 1970a, b) and since this organism can be grown to high cell densities in axenic media, we have studied the DNA associated with isolated pellicles of Tetrahymena.

Tetrahymena pyriformis, strain T, was obtained from the Culture Collection of Algae and Protozoa, Cambridge, and grown as described previously (Flavell & Jones, 1970b). Two-litre cultures were harvested by centrifugation at 500g for 10 min, washed in distilled water and centrifuged at 750g for 15 min. For the preparation of pellicles, cells were grown to late exponential phase (yield approximately 0·8 ml packed cells/100 ml culture) and for preparations of oral plates to early exponential phase (yield approximately 0·4 ml/100 ml).

Ethanol method. Washed cells were suspended in 10 volumes of 40% ethanol in water at −15 °C and stored at −20 °C for 3 h. The cells were then centrifuged at 500g for 10 min at −15 °C, suspended in 0·3 M mannitol, 2 mM EDTA, 10mM Tris-HCl, pH 7·0 and gently homogenized with an Ultra-Turrax tissue homogenizer (Type TP 18/2, Janke and Kunkel K.G., Staufen i. Br., Germany). Pellicles were collected by centrifuging at 500g for 10 min. The white, flocculent pellicular pellet was carefully removed from the surface of a brown pellet of whole cells and washed 4 times with the mannitol medium.

Washed cells were suspended in 0·3 M sucrose, 2 mM EDTA, 10 mM Tris-HCl pH 7·5 (sucrose-EDTA) and homogenized with an Ultra-Turrax. Pellicles were isolated by the same procedures as in the ethanol method.

Oral plates were isolated from early exponential-phase cells by the same procedures as for pellicles. The pellicular membrane seems to be much more labile in such cells and the only intact fragments after homogenization were whole oral plates. These could be isolated in a pure state and phase-contrast microscopy showed them to be free from mitochondria and nuclei.

Mitochondria and nuclei were prepared by methods described previously (Flavell & Jones, 1970 a, b).

Nuclei were suspended in sucrose-EDTA and homogenized or sonicated until extensively disrupted. The homogenate was centrifuged at 300g for 10 min to remove unbroken nuclei and then at 20000g for 10 min to sediment the particulate fraction of the disrupted nuclei. DNA was prepared from the pellets and supernatant fraction.

Pellicles from 15 ml of packed cells were suspended in 50 ml saline-EDTA (0 · 15 M NaCl, 0 · 1 M EDTA, pH 8 · 0) and sodium lauryl sulphate added to a final concentration of 0 25 %. An equal volume of phenol, saturated with saline-EDTA, was added and the mixture shaken gently at 0 °C for 15 min. The mixture was centrifuged to separate the phases and the procedure repeated on the aqueous layer. Dispersion and re-extraction of the material at the phenol-water interface gave no increased yield of DNA. The solution was extensively dialysed against 0 ·05 M sodium phosphate buffer, pH 6·8, ribonuclease added to a concentration of 50 μg/ml, and incubation continued for 30 min at 37 °C. The solution was extracted with phenol saturated with 0·05 M sodium phosphate and applied to a 1 × 1 cm column of hydroxyapatite. Elution with 015 M sodium phosphate was continued until of the effluent fell below 0·01. The DNA was then completely eluted with 0·4 M sodium phosphate and dialysed against several changes of SSC (0·15 M NaCl, 0 ·015 M sodium citrate pH 7·2).

DNA prepared by this method contained negligible RNA or protein and had E₂₆₀/E₂₃₀ > 2 and E₂₆₀/E₂₃₀ = 2 ·3.

Analytical density gradient measurements were performed as described previously (Flavell & Jones, 1970b) using DNA from Micrococcus lysodeikticus as a density standard (ρ = 1731 g cm⁻³).

Pellicles prepared by either the ethanol or sucrose procedures contained few free mitochondria, but in some preparations the large fragments, which were in many cases complete ghosts, contained a small number of mitochondria and other unidentified small particles trapped within the membranous structures. Because of the presence of EDTA in the isolation medium the nuclear membrane was disrupted and no whole nuclei were visible. Preparations of oral plates were free from contamination with other subcellular particles.

The yield of DNA from isolated pellicles was approximately 1–2 μg/ml of packed cells. This may be compared with a yield of 5–10μg of mitochondrial DNA/ml and 250 μg of nuclear DNA/ml of packed cells. Obviously the amount of DNA recovered from pellicles is relatively very small.

When analysed in the caesium chloride gradient, mitochondrial DNA shows a single sharp band at a buoyant density of 1·685 g cm⁻³, and nuclear DNA a major band at 1 ·688 g cm⁻³ and a minor (∼ 2%) at 1·698 g cnr³ (Fig. 1). Pellicle preparations contained all 3 bands in variable relative amounts. In some preparations, the components at 1·685 and U698 g cm⁻³ were enriched relative to a whole cell DNA. Pellicle DNA prepared by either the ethanol or sucrose methods showed similar overall patterns but enrichment of the minor components was more commonly observed with the sucrose procedure. Fig. 2A shows an analysis of an ethanol-prepared sample and Fig. 2B a sucrose preparation.

The effect of treating pellicles with deoxyribonuclease before lysis was tested by incubating them in 0·3 M sucrose, 5 mM MgCl₂, 10 mM Tris-HCl pH 7-0, containing 100 μg/ml deoxyribonuclease. Incubation was continued for 30 min at 0–4 °C. It was expected that a mild enzyme treatment such as this might remove contaminating DNA and thereby enrich any specific pellicular DNA. Wells & Birnstiel (1969) found that when chloroplast preparations were treated with deoxyribonuclease at 0–4 °C, contaminating nuclear DNA was selectively removed, whereas Wolstenholme & Gross (1968) showed that incubation at 37° C removed both chloroplast and nuclear DNA. However, treatment of pellicles merely reduced the yield still further (0·5 μg DNA/ml packed cells) and no enrichment of any satellite component was detected.

The oral plate of Tetrahymena contains a large number of kinetosomes (the postulated site of DNA in the pellicle) and proved to be easy to isolate in a pure form. Several preparations of oral plates were made but no DNA could be isolated from these structures.

The DNA banding at 1·685 g cm⁻³ in the pellicle preparations can be attributed, with a considerable degree of confidence, to mitochondrial DNA, since mitochondria could always be seen in the pellicle preparations.The other 2 components are both present in nuclear DNA but it is somewhat surprising that the band at 1·698 g cm⁻³ should be enriched if random contamination with nuclear DNA was occurring. It appeared possible that this band was a nuclear DNA component contained in a particulate fraction of the nucleus. Following lysis in the presence of EDTA these particles might be trapped within the pellicular envelope as happens with mitochondria. This hypothesis was rendered likely by the observation that the heavy nuclear satellite DNA hybridizes with ribosomal RNA and is therefore the ribosomal DNA of Tetrahymena (Flavell & Jones, unpublished experiments). In other systems the ribosomal DNA has been found to be enriched in nucleolar preparations (McConkey & Hopkins, 1964). To test this possibility nuclei were fractionated as described in Materials and Methods: the procedure approximates to the conditions during pellicle isolation. Nuclei from both strain T and the micronucleate strain A of syngen T were used. Strain A was used since very pure macronuclei can be obtained by a fractionation procedure. The 2 nuclear DNA components have densities 1-685 (major) and 1’693 g cm⁻³ (minor) (Holt & Jones, manuscript in preparation). Table 1 shows the distribution of DNA in the 3 fractions. The heavy satellite band was considerably enriched in the 20000-g pellet from both strain A (Fig. 3 A) and strain T, although interpretation of the latter was made difficult by the broader bands obtained presumably as a result of degradation during isolation. DNA from the 20000-g supernatant gave rise to a broad band at the major nuclear density, with no indication of a satellite component (Fig. 3 B).

If we compare the rather small amount of DNA which seems to be associated with pellicles with the total number of kinetosomes in the pellicle, it is apparent that any specific kinetosomal DNA would necessarily be a molecule of low complexity. It follows that the DNA would renature rapidly in the same way as do mitochondrial and chloroplast DNA. Therefore a specific pellicular DNA concident in buoyant density with nuclear DNA would be detectable by a denaturation-renaturation experiment.

DNA from isolated pellicles was denatured by heating at 100 °C for 10 min and renatured by incubation in 2 × SSC for 2 h at 60 °C. The preparation contained the 3 bands in the relative amounts: 1·685 (20%)> 1·688 (60%) and 1·698 g cm⁻³ (20%). Measurement of E₂₃₀ showed that about of the DNA renatured in the 2-h period. In the caesium chloride gradient (Fig. 2c) only one obviously renatured DNA band is visible at i-686gcm⁻³, identical in density with renatured mitochondrial DNA (Flavell & Jones, 1970b). The remainder of the DNA formed a broad band centred on 1-705 g cm⁻³. No rapidly renaturing band at the density of native nuclear DNA (1 -688 g cm⁻³) could be resolved, and it therefore seems unlikely that pellicular DNA of this density was present in these preparations.

Despite the consistent finding of DNA in pellicle preparations, the evidence presented here indicates that this DNA is, in fact, contamination with other cellular DNA’s. The 3 components resolvable in caesium chloride gradients can be explained as follows:

1. The sharp band at 1·685 g cm⁻³ found in all preparations is identical in buoyant density to mitochondrial DNA and renatures in a similar way. Since mitochondria were regularly observed in pellicle preparations we believe that this band represents mitochondrial DNA.

2. The broad band at 1·688 g cm⁻³ is in all respects similar to the major nuclear component.

3. The component at 1·698 g cm⁻³, while of the same density as the minor nuclear component, was in many preparations present in up to 20% of the total, which may be compared with 1 % found in nuclei. It was possible to produce similar enrichments of this component in subfractions of nuclei and we therefore feel that its presence in pellicle preparations is due to entrapment of this particulate fraction of the nucleus during isolation.

Since we find no rapidly renaturing DNA other than the presumptive mitochondrial DNA, it appears rather unlikely that a specific pellicular DNA is masked by the broad band of the main nuclear component.

Probably the best evidence for the absence of a specific pellicular DNA is the failure to find any DNA in the oral plate preparations. These preparations were very homogeneous and no contamination with other subcellular particles was detected.

It might be argued that the conditions used in these experiments failed to isolate a specific pellicular DNA either because the DNA-containing organelles were not disrupted by the detergent treatment or because the DNA was selectively lost during purification. However, Hufnagel (1969) has shown that sodium lauryl sulphate effectively dissolves basal bodies and we have confirmed this. Further, experiments in which the DNA was extracted (Marmur, 1961) with sodium perchlorate/chloroform/isoamyl alcohol yielded similar DNA, although usually in smaller quantity. We therefore consider the most reasonable interpretation of our results to be that either no specific pellicular DNA exists or that the amount is too small to be detected by these methods.

Since approximately 0·1–0·2 μg of DNA can be detected in the caesium chloride gradient we can estimate that, if there are some 300–500 kinetosomes per cell, a pellicular component of io⁸ Daltons molecular weight should have been detected. It is unlikely that a DNA of less complexity than this could have any genetic significance and, in particular, could hardly be involved in the specification of the complex cortical structures.

Our conclusions are therefore in support of those of Hufnagel (1969) and Pyne (1968) who were unable to demonstrate the presence of a specific pellicle-associated DNA. The only positive indications for a pellicle DNA are based on the autoradiographic studies of Randall & Disbrey (1965) and Smith-Sonneborn & Plaut (1967, 1969) using pellicle fragments isolated following ethanol and non-ionic detergent treatments known to cause lysis of mitochondria. In our view the final decision as to whether a specific pellicular DNA exists can be provided only by in vivo autoradiographic studies or by more sensitive physical methods using isotopic labelling.