Figure 1.
Snapshot of L1- and nested gene-induced TI from UCSC Genome Browser [39].
Schematic representation of (A) intron retention in TRPC4AP (ESTs: DA794245, DB292589) upstream to full-length L1 PA5 element, (B) forced exonization and cryptic polyadenylation in ZNF385D (ESTs: BG194196, BG195414, BG202901, BP276600) upstream to full-length L1 PA7 element, (C) exonization in OPN3 (ESTs: DN996875, DA323545, AA745052 etc) upstream to single exon coding nested gene CHML and (D) intron retention in PGBD2 (EST: DA629789) upstream to putative nc nested gene (EST: AK021482). Arrows above L1 represent the direction of transcription. Red arrows mark the 3′ ends of aberrant transcripts. Although the cryptic polyadenylation is shown only in panel B, it was variably found in either exonization or intron retention products (Figure 8).
Table 1.
Prediction of TI between host and protein-coding nested genes.
Figure 2.
TI effects induced by protein-coding and putative nc nested genes in different human tissues.
Analysis of the transcripts of five nc (panels 3–7) and two protein-coding nested genes (panel 1 and 2). Endogeneous host gene (H), nested gene (N) and host gene transcripts showing exonization and intron retention upstream to the nested genes (TI) were detected by RT-PCR. Intron-containing transcripts in TXNDC12 were further amplified by nested PCR. Names of the host-nested genes and TI transcripts are indicated on the left and length of the products on the right side of panels, respectively. Corresponding exon-intron structures (not in scale) for nested genes are shown on the right of panels. Nested genes are shown with grey boxes, exonization of intronic sequence upstream to nested gene is shown with hatched box with upward diagonals and intron retention with downward diagonals. Exons are shown with white boxes, introns with lines and splicing with diagonal lines. Primers used for each (H, N and TI) amplification are shown in parenthesis on the left side and their numbers refer to the corresponding product shown on the right side of each panel. Their numbers correspond to sequences listed in Table S5. Plus and minus signs represent experiments in the presence or absence of reverse transcriptase (RT). PPIA – positive control; NTera2D1 – human teratocarcinoma cell line.
Figure 3.
Mapping of ESTs used for the prediction of TI in three human genes.
(A) ABCA9 mRNA - NM_080283, ESTs: AI372047, CB960713, CB961243 and L1 PA3 were mapped to its genomic structure containing exons 20 to 26. L1 ASP drives transcription from the opposite strand and produces ESTs CB960713 and CB961243. Intron retention to exon 23 is shown in EST AI372047. (B) NCAM1 mRNA - NM_181351, ESTs: BX432004, BC029119 and L1 PA5 were mapped to its genomic structure containing exons 8 to10. Intron retention to exon 9 is observed in two ESTs (BX432004 and BC029119) one of which is polyadenylated. (C) TXNDC12 mRNA - NM_015913, ESTs: DB460634, DA277685, DA719333, DA735586 and KTI12 mRNA - NM_138417 were mapped to its genomic structure containing exons 1 to 3. Exonization upstream to KTI12 was observed in four ESTs (DB460634, DA277685, DA719333 and DA735586). Genomic DNA is marked with blue line on top. Orange boxes refer to exons, blue lines to introns and arrows indicate the direction of transcription. L1 and KTI12 are shown with orange thick lines. Detailed mapping was carried out with Spidey [69].
Figure 4.
L1-induced transcription decreases SV40 transcription and causes intron retention in ABCA minigene.
(A) Schematic representation of the ABCA minigene and its deletion constructs used in transfection experiments. Splicing of exons observed in different transcripts is shown by diagonal lines. Deletions made in ABCA Fl construct are marked with lines below the scheme. Intron retention to exon 23 is shown by hatched box with downward diagonals. Exonization of intronic sequence upstream to L1 5′ UTR is shown with hatched box with upward diagonals and marked with Ex. In this exon cryptic acceptor splice site (SA) and polyA signal (p(A)) are marked. Additional SV40 polyA signal (marked with lollipop) is located in exon 3. L1 ASP drives transcription from the opposite strand and produces transcript containing exons I, II and III (grey boxes). Its structure is shown on panel A bottom strand. Primers used in RT-PCR are shown with small arrows and their sequences are listed in Table S5. (B) RT-PCR of various transcripts expressed from minigene constructs. Transcripts, containing exons or introns (shown on top of each panel) derived from SV40 promoter and L1 SP/ASP were reverse transcribed and amplified by PCR. The obtained products are shown on the left of each panel and their labeling is according to the scheme A. Size markers (in bp) are shown on the right of each panel. (C) Quantitative detection of various minigene transcripts by RPA. Riboprobes Fl1-23-3 and Int1-23 (lanes1 and 11) are schematically shown above each panel. Fl and different deletion (Δ) constructs (ΔASP = ΔASP292) used in transfection experiments are shown on top of each lane. Protected transcripts (marked with arrows) are schematically shown by boxes and their sizes are given in nucleotides. Dashed lines/boxes show the remaining exon(s) not protected by the riboprobe used. TE – transfection simulated with buffer. Autoradiogram clip on the bottom left shows a 2-fold serial dilution of the probe used in quantitation of transcripts.
Figure 5.
TI effects of L1 SP and ASP, cryptic splice site and polyA signals in ABCA minigene.
(A) Analysis of the L1 ASP- and SP-driven transcripts expressed from various minigene constructs (ΔASP = ΔASP292). Note the large number of endogenous transcripts (endo) with sizes 40–132 nt and 40–89 nt hybridizing to exon I and SP, respectively. (B) Deletion of acceptor splice site increases intron retention. Transcripts expressed from ΔSA containing acceptor splice site deletion are compared to those expressed from Fl and ΔL1 constructs. (C) Analysis of the intron-containing, exonized and polyadenylated transcripts in cytoplasmic and nuclear RNA fractions. Two different types of transcripts were detected with riboprobes shown above panels. Total (Σ), cytoplasmic (cyt) and nuclear (nuc) RNAs were isolated from Fl and ΔL1 transfection experiments. In experiments shown on lanes 1, 4 and 7, half of the amount of RNA was used compared to other experiments. In all experiments (A–C), transcripts detected by RPA are shown by arrows and arrowheads (faint bands) and they are schematically represented according to the scheme shown in Figure 4A. The numbers mark their sizes in nucleotides. When both, endogenous and minigene transcripts were detected, endo is shown in parenthesis. Quantitative estimation of transcripts was carried out by comparison with 2-fold serial dilution of the riboprobe shown in panel B bottom left.
Figure 6.
L1-induced transcription interferes with exon skipping and intron retention in NCAM minigene.
(A) Schematic representation of the NCAM minigene and its deletion constructs used in transfection experiments. Deletions made in NCAM Fl construct are marked with lines below the scheme. Intron retention to exon 9 is shown by hatched box. Potential transcription from L1 ASP (although not proved by experiments, see text for details) is shown from the opposite strand. For other details see Figure 4 legend. (B) RT-PCR of various transcripts expressed from minigene constructs. Transcripts, containing exons or introns (shown on top of each panel) derived from SV40 promoter and L1 SP were detected by RT-PCR. The obtained products are shown on the left of each panel and their labeling is according to scheme A. (C) Quantitative detection of various minigene transcripts by RPA. Riboprobes Fl1-9-3, Int1-9 and L1SP-3 (lanes 2, 8 and 14) are schematically shown above each panel. Fl and different deletion (Δ) constructs used in transfection experiments are shown on top of each lane. Protected transcripts (marked with arrows) are schematically shown by boxes and their sizes are given in nucleotides. Dashed lines/boxes show the remaining exon(s) not protected by the riboprobe used. TE – transfection simulated with buffer. Autoradiogram clip on the bottom left shows a 2-fold serial dilution of the probe used in quantitation of transcripts. (D) Distribution of intron-containing transcripts between cytoplasm and nucleus. Total (Σ), cytoplasmic (cyt) and nuclear (nuc) RNAs were isolated from Fl and ΔL1 transfection experiments and probed with intron-containing probe.
Figure 7.
KTI12-induced transcription interferes with exonization and intron retention in its upstream region.
(A) Schematic representation of the TX-KTI minigene and its deletion constructs used in transfection experiments. Deletions made in TX-KTI Fl construct are marked with lines below the scheme. Splicing of exons observed in different transcripts is shown by diagonal lines. Exonization of intronic sequence upstream to KTI12 is shown by hatched box and marked with Ex. Arrows indicate the direction of transcription of SV40 and KTI12. (B) RT-PCR of various transcripts expressed from minigene constructs. Transcripts, derived from SV40 and KTI12 promoters, containing exons or introns (shown on top of each panel) were detected by RT-PCR. The obtained products are shown on the left of each panel and their labeling is according to the scheme A. (C) Quantitative detection of various minigene transcripts by RPA. Riboprobes are schematically shown above each panel. Fl and different deletion (Δ) constructs used in transfection experiments are shown on top of each lane. Protected transcripts (marked with arrows) are schematically shown by boxes and their sizes are given in nucleotides. Dashed lines/boxes show the remaining exon(s) not protected by the riboprobe used. The smear of partially protected fragments (shown on lanes 10 and 15) most likely corresponds to prematurely terminated transcripts. TE - transfection simulated with buffer. Autoradiogram clip on the bottom right shows a 2-fold serial dilution of the probe used in quantitation of transcripts. (D) Detection of various TI transcripts. Intron-containing riboprobes used are shown above each panel. Protected transcripts are schematically shown as in (C).
Figure 8.
TI effects induced by human L1 retrotransposon and nested gene.
(A) Tandemly arranged intronic L1 interferes with the host gene transcriptional elongation by forcing intron retention, exonization and cryptic polyadenylation. The TI effect of L1 depends on its SP and ASP activity, cryptic splice sites and polyA signals located upstream to L1. (B) A tandemly arranged nested gene interferes with the host gene transcription by causing exonization and intron retention in its upstream region. Pol II complexes are shown with ellipses (host in blue, nested in green and L1 SP and ASP in yellow and red, respectively) and their direction of transcription with arrows. Exons are displayed with boxes and introns with lines. Splicing is shown by diagonal lines. Intron retention and exonization are shown with hatched box with downward and upward diagonals, respectively. Frequency (%) of exonization, intron retention and polyadenylation induced by L1 5′ UTR and protein-coding and nc nested genes was determined from the data shown in Table S1 and Table S3.