Two Types of scnRNAs
(A) A Tetrahymena cell contains a macronucleus (MAC) and a micronucleus (MIC). During vegetative growth, both the MAC and the MIC divide and segregate to daughter cells. Mixing starved cells of different mating types induces conjugation (i). The MICs undergo meiosis (ii), and one of the selected products divides mitotically to form two pronuclei (iii). One of the pronuclei crosses the conjugation bridge (iv) and fuses with the stationary pronucleus to produce the zygotic nucleus (v), which then divides twice (vi) to form two new MACs and two MICs (vii). The parental MAC is degraded, and the pair is dissolved (viii). The exconjugants resume vegetative growth when the nutrient supply is restored (ix). The approximate time when each event occurs is indicated (hpm, hours post-mixing).
(B) 1,464 MIC genome supercontigs (SCs, blue bars) were ordered by their lengths (longest to shortest) and concatenated. Normalized numbers (reads per kilobase per million reads [RPKM]) of sequenced 26- to 32-nt RNAs from WT cells at the indicated time points that map uniquely to the MIC genome are shown as histograms with 50-kb bins. The densities of IESs and mappable (unique) sequences are also shown. The drop in IES density in the region containing very short SCs is probably because these SCs are shorter than most of the IESs, and the prediction of IESs from them failed. The regions enlarged in (C and D) are marked with green lines. Longer (1–50) and shorter (51–1,464) MIC SCs represent B- and A-regions of the MIC genome, respectively.
(C and D) Small RNA expression from the indicated 300-kb windows (shown and analyzed as in B, except with 100-nt bins). Colored boxes indicate the positions of IESs (magenta, type A; sky blue, type B; see for the IES classification). In (D), the arrows mark MAC-destined regions that are the origins of Early-scnRNAs that accumulated prominently at early stages (3 hpm) but were degraded later.

Late-scnRNAs Interact with Twi1p and Twi11p
(A) Expression of the Argonaute genes in vegetatively growing (l, m, and h indicate low, medium, and high density, respectively), starved (numbers indicate hours after removal of nutrients), and conjugating (numbers indicate hpm) cells based on a microarray analysis.
(B) Twi1p and Twi11p from wild-type cells at the indicated time points of conjugation were detected by western blotting using the indicated antibodies.
(C) Different transgenic and gene-knockout strategies. In the vegetative and early- to mid-stage conjugating cells, the parental MACs (yellow) contribute all (maternal) gene expression, whereas the new MACs, which are formed from the MIC (green), provide zygotic gene expression at late conjugation.
(D and E) Proteins from cells expressing FLAG-HA-Twi1p from the maternal TWI1 loci by MAC replacement (D) or cells expressing HA-Twi11p from the zygotic TWI11 loci by MIC replacement (E) at the indicated time points of conjugation were detected by western blotting.
(F) HA-Twi11p was immunopurified using an anti-HA antibody (α-HA-IP) from the HA-TWI11 conjugating culture at 10.5 hpm. As a negative control, wild-type cells (WT) were used for similar immunopurification. Early-scnRNAs were immunopurified with Twi1p by an anti-Twi1p antibody (α-Twi1p-IP) from the wild-type conjugating culture at 3 hpm. RNAs were separated on denaturing gels and stained using a nucleic-acid-specific dye. Arrowhead indicates scnRNAs.
(G–Q) 26- to 32-nt RNAs from the indicated immunopurified RNAs or strains were analyzed as in B. The 300-kb window on the right is marked with a green line.
See also .

Three Types of IESs
(A) Classification of IESs according to the expression of Early- and Late-scnRNAs.
(B) Localization of different types of IESs, transposons (TEs), and coding sequences (CDSs) in the MIC genome are shown in a histogram with 50-kb bins.
(C and D) Mean lengths (C) and GC contents (D) of IESs in different IES classes.
(E) Distributions of TE-related sequences among MIC genome components (left) and IES types (right). All possible 25-mers from the MIC genome sequences were classified as sequences that were complementary to only IESs (red), only MAC-destined sequences (MDSs, blue), or both (yellow). All possible 25-mers from the total IESs were classified as sequences that were complementary to only Type-A IESs (magenta), only Type-B IESs (sky blue), or both (purple). The fraction of TE-derived 25-nt sequences complementary to these DNA classes was calculated.
(F–I) Analysis of DNA elimination efficiency. The retention indexes (RIs) of individual IESs in the purified new MACs of the indicated strains at 36 hpm were plotted. IESs are ordered according to their LEIs (on the x axis). The red lines indicate RI = 1 (no DNA elimination).

trans Recognition of IESs
(A–C) Top two panels: normalized numbers (RPM) of sequenced 26- to 32-nt RNAs from wild-type cells at 3 hpm mapping to the three representative Type-B IES loci (10 kb) are shown as histograms with 50-nt bins. For the top histograms, only the numbers of sequences uniquely mapping to the MIC genome are shown (unique mappers). The middle histograms show the numbers of sequence reads mapping to each position within the loci divided by the total numbers of sites in the entire MIC genome to which the sequence reads map (weighted). Arrows indicate regions to which Early-scnRNAs map. Sky-blue boxes represent the IESs. Bottom panels (repeats): all possible 25-mers from the entire MIC genome sequence (gray), from Type-A IESs (magenta), or from Type-B IESs (sky blue) were mapped to the three Type-B IES regions, and their frequencies of occurrence are shown as histograms with 50-nt bins.
(D) All possible 25-mers were extracted from the indicated IESs, and their frequencies of occurrence (hits per kilobase [HPK]) on the MIC genome are shown as histograms with 50-kb bins. The locations of the IESs are marked with red dots. The density of the Type-A IESs is shown at the top.
(E–G) Three representative Type-A IES (magenta boxes) loci were analyzed as in (A)–(C).
See also .

Role of A-Repeats in DNA Elimination
(A) DNA elimination assay of R-IES. R-IES (a Type-B IES) and its flanking MDS regions was inserted into the extra-chromosomal vector pD5H8 and introduced into the developing MAC. Progeny lines were established and elimination of R-IES on the vector was analyzed by PCR using primers complementary to the vector (arrows). The PCR results are shown on the bottom. The arrowhead and double arrowhead indicate the expected position of the PCR products from R-IES loci on the vector without or with R-IES elimination, respectively.
(B) DNA elimination assay of a pseudo-IES. Most of the R-IES in pD5H8 shown in (A) was replaced by a GFP-coding sequence and analyzed as in (A).
(C) DNA elimination assays assessing the DNA elimination-inducing activities of IES fragments. A fragment of an IES of interest (red boxes with numbers) was inserted into the middle of the pseudo-IES shown in (B) and tested to see if the fragment restored DNA elimination of the pseudo-IES. The progeny cell lines showing any deletion at the pseudo-IES loci (marked with asterisks) were counted. The arrowheads indicate PCR products from the pseudo-IES loci on the vector without any DNA elimination. The double arrowheads show the expected positions of PCR products from the pseudo-IES loci on the vector following DNA elimination exactly at the borders of the original R-IES.

Co-deletion
(A–C) Normalized and weighted numbers of sequenced total 26- to 32-nt small RNAs from wild-type cells at indicated time points mapping to the representative Type-B IES loci are shown as histograms with 50-nt bins. Sky-blue boxes indicate IESs.
(D and E) Schematic drawings of the co-deletion (coDel) experiment (D) and of the hypothetical actions of scnRNAs in coDel (E). See the text for details.
(F–M) Indicated IESs and target loci were cloned into the extra-chromosomal vector pD5H8 and introduced into the new MACs of conjugating wild-type (F–L) or TWI11 MIC-KO (M) cells. Deletions at the endogenous target loci in progeny cells were analyzed by PCR as in (D). For the “w/o target” experiment (H), the TTHERM_00079350 locus was analyzed. The number of progeny lines showing any deletions at the endogenous target loci was determined. The arrowheads and brackets indicate PCR products from the target loci without or with deletions, respectively. The double arrowheads show the expected positions of PCR products from the target loci with deletions exactly corresponding to the target sequences.
See also .

A trans-Recognition Network in IES Recognition
(A) Results of a simulation of accidental loss of Early-scnRNA expression. See the text for details. The average fractions of 1° and 2° IESs in each condition, with three different sets of randomly chosen seed Type-A IESs (47 and 469 Type-A IESs for 1% and 10% seeds, respectively) are shown.
(B) A model for IES recognition. Early-scnRNAs are expressed from Type-A IESs (magenta boxes) and their flanking sequences in the MIC (i), and the latter are degraded in the parental MAC (ii). In the new MAC, Early-scnRNAs recognize the Type-A IESs from which they are derived (iii) as well as other Type-A and Type-B (sky-blue boxes) IESs in trans (iv) through A-repeats (filled orange boxes) to trigger Late-scnRNA production (v). In an IES, regions producing Late-scnRNAs spread in cis (vi). Late-scnRNAs further trans-recognize other IESs (vii).