The SUMO pathway is developmentally regulated and required for programmed DNA elimination in Paramecium tetraurelia

Abstract

Extensive genome-wide remodeling occurs during the formation of the somatic macronuclei from the germ line micronuclei in ciliated protozoa. This process is limited to sexual reproduction and includes DNA amplification, chromosome fragmentation, and the elimination of internal segments of DNA. Our efforts to define the pathways regulating these events revealed a gene encoding a homologue of ubiquitin activating enzyme 2 (UBA2) that is upregulated at the onset of macronuclear development in Paramecium tetraurelia. Uba2 enzymes are known to activate the protein called small ubiquitin-related modifier (SUMO) that is covalently attached to target proteins. Consistent with this relationship, Northern analysis showed increased abundance of SUMO transcripts during sexual reproduction in Paramecium. RNA interference (RNAi) against UBA2 or SUMO during vegetative growth had little effect on cell survival or fission rates. In contrast, RNAi of mating cells resulted in failure to form a functional macronucleus. Despite normal amplification of the genome, excision of internal eliminated sequences was completely blocked. Additional experiments showed that the homologous UBA2 and SUMO genes in Tetrahymena thermophila are also upregulated during conjugation. These results provide evidence for the developmental regulation of the SUMO pathway in ciliates and suggest a key role for the pathway in controlling genome remodeling.

FIG. 1.

Expression of UBA2 and SUMO in Paramecium. (A) Northern blot of total RNA (20 μg per lane) from Paramecium probed sequentially, first with Paramecium UBA2 and then with a mixture of SUMOI, SUMOII, and SUMOIII genes. Ethidium bromide staining of rRNA was used as a loading control. (B) Protein levels of Uba2-GFP (filled circle) and GFP-SUMOI (open circle) were estimated by measuring fluorescence intensity (in arbitrary units) in vegetative and mated cells. Expression of each GFP fusion gene was driven by its own upstream regions. Examples of pictures are shown in Fig. 2C and D. Vertical bars represent standard deviation.

FIG. 2.

Cellular localization of GFP fusion proteins. (A and B) Fusion protein of GFP to the C terminus of Uba2p (A) or N terminus of SUMO I (B) in autogamous cells expressed by their own promoters. The nuclear events in these cells correspond to approximately 12 to 20 h in conjugating cells. Fluorescent images are projections of optical sections obtained by confocal microscopy. The GFP fluorescence localized to the new macronuclei, which show weaker propidium iodide staining (DNA) than the old macronuclear fragments. Pictures on the right are ×3 magnifications of a new macronucleus. The bar is 15 μm for the left pictures and 5 μm for ×3 magnification on the right. (C and D) GFP fluorescence from Uba2-GFP (C) and GFP-SUMOI (D) expressed by their own upstream regions in mated cells. Pictures for GFP fluorescence were taken at the same exposure under a conventional fluorescent microscope. The bar in panel C corresponds to 20 μm for both panels C and D.

FIG. 3.

Cytological phenotypes of RNAi-treated exconjugants. (A) Northern blot of RNAi-treated autogamous cells probed sequentially with UBA2 and T2c (alpha-tubulin [α-tub]). (B) Representative DAPI-stained cells of normal and defective cytological phenotypes observed in RNAi treatments. Each picture contains a single cell with or without the new macronucleus (arrowheads) and old macronuclear fragments. (C) Phenotypic classes of the left three were assigned as indicated below the pictures and are plotted in the graphs. The right two panels show pictures of cells containing round, immature macronucleus characteristic of UBA2 and SUMO RNAi. Mac, macronucleus. (D) DNA content in the new macronucleus of exconjugants. Exconjugants were stained by propidium iodide, and the fluorescent signal in the nuclei was measured. Vertical bars represent standard error.

FIG. 4.

Inhibition of excision of six IESs in the A-51 allele by RNAi of UBA2 and SUMO. (A) A partial map of the micronuclear version of the A-51 allele showing locations of IESs (boxes), SspI recognition sites (arrow heads), and sizes of fragments generated by SspI digestion. Positions of expected PCR products (pp1 to pp4) and probes for Southern hybridization are also indicated. Arrows show positions of the primers relative to IESs not drawn to scale. (B) Whole-cell semiquantitative PCRs of RNAi-treated exconjugants using one primer in the macronuclear sequence and the other primer inside the IESs. Each lane represents whole-cell PCR products taken at 2-h intervals from 6 to 22 h after induction of conjugation (conj.). (A) The predicted PCR products correspond to pp1 to pp4 in panel A. Due to excision of smaller IESs during rearrangement, two bands are expected for each primer set. (C) SspI-digested total genomic Southern blots of RNAi-treated exautogamous cells probed either with a HincII-PstI fragment or IES4578, as indicated in panel A. Total DNA (∼10 μg) was isolated from an exautogamous cell culture when about 50% of control cells (RNAi using empty vector) had undergone the first cell division. Most IESs in the micronuclear version of the A-51 allele contains SspI sites (single arrow heads in panel A), while only one site is present in the macronuclear-destined sequence of the A-51 allele (the double arrow head in panel A). Thus, for probe HincII-PstI, 1.3- and 2.0-kb fragments are expected for unprocessed DNA, while a 5.7-kb fragment is expected for the processed DNA, including abundant old macronuclear DNA in exconjugants. Probe IES4578 contains only the IES sequence and detects 0.5-kb fragments if the IES is present at high levels in exautogamous cells.

FIG. 5.

RNAi of UBA2 and SUMO did not affect vegetative growth of Paramecium. (A) RNAi effects for three continuous days were monitored in 12 sublines with extrachromosomal expression of GFP-Uba2p. Upon UBA2 RNAi, GFP signal was effectively reduced (solid circles; bars on the circles show 95% confidence), while cell division rates in both vector and UBA2 RNAi-treated cells were unaffected (vertical bars show standard deviation). (B) Daily isolation lines (12 to 24 lines) of wild-type (GFP-nontransformed) cells were treated with RNAi for three continuous days, and survival of the cell line and fissions per day were counted.

FIG. 6.

Northern blot of Tetrahymena total RNA (20 μg per lane) from several developmental time points probed with Tetrahymena UBA2 (top panel) or SUMO (bottom panel).