Search

During their proliferation and the host’s concomitant attempts to suppress it, LINE-1 (L1) retrotransposons give rise to a collection of heterogeneous ribonucleoproteins (RNPs); their protein and RNA compositions remain poorly defined. The constituents of L1-associated macromolecules can differ depending on numerous factors, including, for example, position within the L1 life cycle, whether the macromolecule is productive or under suppression, and the cell type within which the proliferation is occurring. This chapter describes techniques that aid the capture and characterization of protein and RNA components of L1 macromolecules from tissues that natively express them. The protocols described have been applied to embryonal carcinoma cell lines that are popular model systems for L1 molecular biology (e.g., N2102Ep, NTERA-2, and PA-1 cells), as well as colorectal cancer tissues. N2102Ep cells are given as the use case for this chapter; the protocols should be applicable to essentially any tissue exhibiting endogenous L1 expression with minor modifications.

Additional file 1: Figure S1. (A) LINE-1 RNA in the input for each cell LINE-1. (B) LINE-1 RNA after ORF1p-IP enrichment.

Additional file 4: Figure S4. qPCR quantification of p-body and ORF1 IP RNA after siRNA knockdown of LINE-1. (A) qPCR of top ORF1 IP and p-body genes after 48 h siRNA LINE-1 knockdown. Normalized using RPL19. (B) Western blot of 48 h siRNA LINE-1 knockdown used for qPCR in (A). Demonstrates knockdown of ORF1p protein.

Additional file 3: Figure S3. Overlap between transcripts enriched in ORF1p IP and in cytoplasmic granules. (A) Overlap between ORF1p IP enrichment and p-body enrichment [55]. A 5% FDR cutoff was used. (B) As in A, but comparing ORF1p IP to SG enrichment [54]. (C) Overlap between transcripts enriched in ORF1p IP and those whose expression is positively correlated with LINE-1 RNA in TCGA prostate cancer samples (i.e. these genes are more highly expressed in tumors that express more LINE-1). (D) Overlap between transcripts enriched in p-bodies and those correlated with LINE-1 RNA in TCGA prostate cancer.

Additional file 1: Figure S1. CLUSTAL Alignments of ORF2 protein sequences. ORF2p protein sequences were obtained from the L1Base database of reference L1Hs sequences. (A) Alignment of 146 full-length L1 sequences. (B) Alignment of 107 ORF2-intact L1 sequences. In the center tiles, black bars indicate amino acid positions where the L1 in that row differs from the CLUSTAL alignment consensus sequence. The ‘% agreement,’ or identity, at each amino acid position is quantified below the center tiles. On the right, the number of amino acid changes of a particular L1 compared to the immunogen, L1RP, is quantified.

Additional file 5: Figure S3. Coomassie G-250 stained gel plugs used for in-gel digestion followed by MS. A panel is shown for every replicate included in the LFQ-MS analysis. (A) Tumor A (Krukenberg Carcinoma, Ovary) was subjected to two independent affinity isolations with different parameters (see Methods). Each isolation included three replicates using anti-ORF1p-coupled affinity medium to capture ORF1p from the tumor extracts (Tumor A-1 to A-6), and three replicates using mouse IgG-coupled affinity medium to sample non-specific background from the same extracts (mIgG A Ctrl-1 to Ctrl-6). (B) Tumor B (Metastatic Rectal Adenocarcinoma, Liver): including three replicates using anti-ORF1p-coupled affinity medium to capture ORF1p from the tumor extracts (Tumor B-1 to B-3), three replicates using mouse IgG-coupled affinity medium to sample non-specific background from the same extracts (mIgG B Ctrl-1 to Ctrl-6), and three replicates using anti-ORF1p-coupled affinity medium to capture ORF1p from matched normal tissue extracts (Normal B-1 to B-3). (C) Tumor C (Adenocarcinoma, Colon): including three replicates using anti-ORF1p-coupled affinity medium to capture ORF1p from the tumor extracts (Tumor C-1 to C-3), three replicates using mouse IgG-coupled affinity medium to sample non-specific background from the same extracts (mIgG C Ctrl-1 to Ctrl-6), and three replicates using anti-ORF1p-coupled affinity medium to capture ORF1p from matched normal tissue extracts (Normal C-1 to C-3).

Additional file 3: Figure S4. Co-IP/Western blot. Three different segments of Tumor D were used as starting material for anti-ORF1p affinity isolations (α-ORF1p T1–3), including a mock-capture control using mouse IgG affinity medium with tumor extracts (mIgG T1), and matched normal tissue with anti-ORF1p affinity medium (α-ORF1p N). Co-IP of ORF1p/2p ectopically expressed from pMT302 in HEK-293TLD is provided as a comparative positive control. All co-IPs used 100 mg cells or tissues as input. 100% of the co-IP elutions done using patient tissues were analyzed; in contrast, fractions (labeled) of the co-IP from pMT302 in HEK-293TLD were analyzed. ORF1p yields from Tumor D were comparable to those obtained from 1/5th – 1/10th of a co-IP from pMT302/HEK-293TLD. However, while ORF2p signal is clearly detectable in 1/5th and closer to the baseline (but still eminently detectable) in 1/10th of a pMT302/HEK-293TLD co-IP, no ORF2p signal was observed in tumor D co-IPs.

Additional file 4: Figure S2. Western blot α-ORF1p titer to detect endogenous ORF1p in clarified cell extracts. The concentration of α-ORF1p used is given along the top; the source of each cell extract is given below that, and each accords to Fig. 2e. The quantity of clarified cell extracts used, in μg total protein, follows below each extract source. I: clarified extract used as an input for α-ORF1p affinity capture; S: immuno-depleted extracts after incubation with α-ORF1p affinity medium. (Left blot image) 1x α-ORF1p concentration - ORF1p signal is observed in with ectopic expression (pMT302) and at just above background in PA-1. α-UPF1 provided as a loading control (NYU1.1B6, 1:1000 [79]). (Right blot image) 5x α-ORF1p concentration - ORF1p signal is observed in all cases except HeLa Kyoto. An increase in non-specific signal is also observed elsewhere on the blot. α-PCNA is provided as a loading control (Santa Cruz Biotechnology, Inc. #sc-56; 1:1000).

Single cell sorting by FACS. Suspension cells were labelled with PI. SSC-width and height, FSC-width and height, and PI were used to gate a subset of live single cell population, and to make sure that only live single cells were sorted.

Comparison of sensitivity for all gold standard L1Hs and only heterozygous non-reference insertions. Black, bulk DNA TIPseq; red, MDA whole genome amplification followed by restriction enzyme digestion and ligation with vectorette adaptors (MDA-D); orange, MDA whole genome amplification followed by end repair, dA tailing and ligation with dT vectorette adaptor (MDA-T); dark blue, MALBAC whole genome amplification followed by ligation with dT vectorette adaptor (MALBAC-T). Diamond, bulk DNA TIPseq; circle, whole genome amplification using random hexamers only (R); square, whole genome amplification using random hexamers and L1 primer (RL).

Supplementary methods. Detailed description of RepProfile prior and simplifications for EM. Also included: rationale for choosing to focus on FB4_DM, DNAREP1_DM and PROTOP. (PDF 133 kb)

Supplementary methods. Detailed description of RepProfile prior and simplifications for EM. Also included: rationale for choosing to focus on FB4_DM, DNAREP1_DM and PROTOP. (PDF 133 kb)

Supplementary results. Descriptions of TE families with 1000 or more edit sites. (PDF 32 kb)

rdgA FD4_DM. RNAstructure prediction for rdgA FD4_DM. The FD4_DM element forms a dsRNA structure that is too long to fit on a figure. (PDF 137 kb)