Gamma radiation increases endonuclease-dependent L1 retrotransposition in a cultured cell assay

Abstract

Long Interspersed Elements (LINE-1s, L1s) are the most active mobile elements in the human genome and account for a significant fraction of its mass. The propagation of L1 in the human genome requires disruption and repair of DNA at the site of integration. As Barbara McClintock first hypothesized, genotoxic stress may contribute to the mobilization of transposable elements, and conversely, element mobility may contribute to genotoxic stress. We tested the ability of genotoxic agents to increase L1 retrotransposition in a cultured cell assay. We observed that cells exposed to gamma radiation exhibited increased levels of L1 retrotransposition. The L1 retrotransposition frequency was proportional to the number of phosphorylated H2AX foci, an indicator of genotoxic stress. To explore the role of the L1 endonuclease in this context, endonuclease-deficient tagged L1 constructs were produced and tested for their activity in irradiated cells. The activity of the endonuclease-deficient L1 was very low in irradiated cells, suggesting that most L1 insertions in irradiated cells still use the L1 endonuclease. Consistent with this interpretation, DNA sequences that flank L1 insertions in irradiated cells harbored target site duplications. These results suggest that increased L1 retrotransposition in irradiated cells is endonuclease dependent. The mobilization of L1 in irradiated cells potentially contributes to genomic instability and could be a driving force for secondary mutations in patients undergoing radiation therapy.

Figure 1

Gamma radiation increases the L1 retrotransposition frequency. (A) The 143B human osteosarcoma cells transfected with L1-EGFP were exposed to gamma radiation and subjected to puromycin selection for the presence of plasmid. The percentage of cells expressing EGFP was measured by flow cytometry. (B) Percentage of EGFP+ cells on day 12 from three independent L1-EGFP transient transfections (triangles, circles and squares) following 0, 2 or 4 Gy of gamma irradiation and puromycin selection. ^*P < 0.0005, ^**P < 0.0001, as compared to 0 Gy by two tailed Student's t-test. A functionally inactive L1 was also tested and had undetectable activity at each dose of irradiation (data not shown).

Figure 2

Potential mechanisms of gamma radiation-induced retrotransposition. Left arrow: gamma radiation-induced double strand breaks could serve as substrates for L1 insertion. Insertions at pre-formed DNA breaks may not require L1 endonuclease; genomic DNA flanking such insertions is expected to lack target site duplications and contain deletions. Right arrow: gamma irradiation may make the host environment more amenable to retrotransposition by upregulating cofactors or downregulating repressors for endonuclease-dependent retrotransposition. An alternative pathway (not shown) is that L1 inserts near DSBs in irradiated cells, but still uses its endonuclease to cleave and/or process DNA at the integration site. This alternative is discussed in the text.

Figure 3

Gamma irradiation and calicheamicin γ1 generate comparable numbers of γH2AX foci. 143B cells were permeabilized and fixed 6 and 24 h after exposure to gamma irradiation (54 and 72 h post-transfection). Double strand breaks were detected by staining for γH2AX. Nuclei were visualized by DAPI staining. (A) Fluorescent images showing γH2AX foci at 6 and 24 h post-irradiation. All images are 400×. (B) Average numbers of repair foci per cell 6 h after exposure to mock, 4 Gy gamma radiation or 5 and 10 pM of calicheamicin γ1. The bar graphs show the mean number of γH2AX foci per cell determined from three separate assessments of at least 100 cells in different regions of the coverslip. Error bars indicate standard deviation.

Figure 4

Endonuclease-independent L1 retrotransposition in CHO-K1 cells subjected to gamma radiation. (A) An endonuclease-deficient L1 was generated by swapping in a portion of ORF2 from L1.3 containing a point mutation at the endonuclease active site (see Materials and Methods). A control endonuclease-competent chimeric L1 element was also generated. (B) CHO-K1 cells were transfected with endonuclease-competent L1-EGFP and retrotransposition measured as shown in Figure 1. n = 9 (three independent transfections measured in triplicate) for each bar. ^*P < 0.001 by two tailed Student's t-test. (C) CHO-K1 cells were transfected with endonuclease-deficient L1-EGFP and retrotransposition was measured. Total EGFP+ events per gated (live) events: 0 Gy: 43/2.3 × 10⁶, 4 Gy: 154/2.2 × 10⁶. The difference in retrotransposition frequency between the irradiated and unirradiated endonuclease-deficient L1s is not significant. A retrotransposition-incompetent L1 did not exhibit detectable retrotransposition with or without irradiation.

Figure 5

L1 insertions from irradiated clones have endonuclease-dependent features. Single L1-EGFP transfected 143B cells expressing EGFP were isolated by flow cytometry and expanded. DNA was extracted and the 3′ genomic flank amplified and sequenced by suppression PCR or inverse PCR (see Materials and Methods). The 5′ flank was then amplified and sequenced using primers designed from the human genome database. Hallmarks of endonuclease-dependent L1 insertion include 7–20 bp target site duplications (TSDs), AT rich consensus target sites and poly-A tails. Dark gray boxes denote TSDs. TSD sequences are displayed beneath each dark gray box. Numbers represent map positions in L1-EGFP; a full length insertion (including the spliced EGFP cassette) is 7814 bp long. Poly-A tail length is given as the subscripted number next to A/T. Chromosome insertion location is given in the 3′ flank. (A) L1 insertion flanks from an unirradiated cell. (B) Insertion flanks recovered following 4 Gy irradiation resemble most endonuclease-dependent genomic L1 insertions. (C) An endonuclease-deficient L1 insertion in a CHO-K1 cell has a deletion at the site of insertion, lacks target site duplications and has 5′ transduced sequence.