Activation of Human Long Interspersed Nuclear Element 1 Retrotransposition by Benzo(a)pyrene, an Ubiquitous Environmental Carcinogen Free

Long interspersed nuclear elements [LINE-1 (L1)] are abundant retrotransposons in mammalian genomes that remain silent under most conditions. Cellular stress signals activate L1, but the molecular mechanisms controlling L1 activation remain unclear. Evidence is presented here that benzo(a)pyrene (BaP), an environmental hydrocarbon metabolized by mammalian cytochrome P450s to reactive carcinogenic intermediates, increases L1 retrotransposition in HeLa cells. Increased retrotransposition is mediated by up-regulation of L1 RNA levels, increased L1 cDNA synthesis, and stable genomic integration. Activation of L1 is dependent on the ability of BaP to cause DNA damage because it is absent in HeLa cells challenged with nongenotoxic hydrocarbon carcinogens. Thus, the mutations and genomic instability observed in human populations exposed to genotoxic environmental hydrocarbons may involve epigenetic activation of mobile elements dispersed throughout the human genome. (Cancer Res 2006; 66(5): 2616-20)

Long interspersed nuclear elements [LINE-1 (L1)] are abundant retrotransposon sequences that occupy ∼17% of the human genome (1, 2). Although the majority of L1 elements are truncated at their 5′-end, ∼100 L1 elements in human, 261 in mouse, and 163 in the rat genome are retrotransposition competent (3–5). A functional human L1 element is ∼6 kb long and consists of a 5′-untranslated region (UTR) with an internal promoter, two open reading frames (ORF1 and ORF2), and a 3′-UTR terminating in polyadenylic acid tail (6). The functional significance of ORF1 is not clear, although purified ORF1 protein is a high-affinity, non-sequence-specific RNA-binding protein (7). ORF2 contains three domains critical for L1 propagation: endonuclease (8), reverse transcriptase (9), and a 3′-terminal zinc finger-like domain (10).

L1 retrotransposition requires the transcription of L1 RNA, its transport to the cytoplasm, and translation of ORF1 and ORF2. These proteins preferentially associate with their encoding transcript to form a ribonucleoprotein particle (11). After transport to the nucleus, L1-encoded endonuclease cleaves genomic DNA at degenerate consensus target sequence, TTTT/A, and variants of that sequence (8, 12, 13). Cleaved DNA provides a 3′-hydroxyl group, which serves as a primer for reverse transcription of L1 RNA by L1 reverse transcriptase (14). After second-strand synthesis, L1 DNA is joined to genomic DNA.

The cellular mechanisms of L1 activation have not yet been fully elucidated. Evidence has been presented that L1 expression in vivo is limited almost exclusively to germ-line and embryonic cells (15–18). Recently, an immunohistochemical study revealed coexpression of ORF1 and ORF2 of L1 in pre-spermatogonia of fetal cells, in germ cells and adult testis, and in distinct somatic cell types, such as Leydig, Sertoli, and vascular endothelial cells (19). Factors that determine cellular specificity of L1 expression have not been identified.

The cellular signals that mediate L1 activation in somatic cells may involve stress signaling through DNA damage. Earlier studies revealed that on exposure to UV light and ionizing radiation L1 elements are dramatically induced in cultured rat chloroleukemia cells (20). In addition, DNA-damaging agents, such as cisplatin, etoposide, and γ-irradiation, substantially induce endogenous reverse transcriptase activity in murine cells (21). Therefore, the activation of L1 sequences may lead to higher L1 retrotransposition rates in target cells. In addition to its own mobility, L1 is also involved in processed pseudogene formation (22), exon shuffling (23), and double-strand break repair (13). The involvement of the human L1 in AluI transposition has been also shown (24–26).

The molecular mechanisms that mediate L1 activation by diverse stress signals are not well understood. We have shown previously that benzo(a)pyrene (BaP), a polycyclic aromatic hydrocarbon carcinogen, activates L1 expression in various cell types (27–29). L1 activation by BaP requires cytochrome P450–catalyed oxidation of the parent hydrocarbon to oxygenated intermediates that adduct DNA (28).

Evidence is presented here that genotoxic damage by BaP is associated with L1 retrotransposition in HeLa cells. Increased retrotransposition is mediated by up-regulation of L1 RNA levels, increased L1 cDNA synthesis, and stable genomic integration into the cells. Activation of L1 is dependent on the ability of BaP to cause DNA damage because it is absent in HeLa cells challenged with nongenotoxic aromatic hydrocarbon carcinogens. Thus, the mutations and genomic instability observed in human populations exposed to genotoxic environmental hydrocarbons may involve activation of mobile elements dispersed throughout the human genome.

Cell culture and stable transfection. HeLa cell cultures were grown in DMEM (Mediatech, Herdon, VA) containing glutamine, supplemented with 10% fetal bovine serum (Atlanta Biologicals, Atlanta, GA) and antibiotics (Invitrogen, Carlsbad, CA), and maintained at 37°C in 5% CO₂. Cultures were grown to <50% confluence and transfected with JM101RP and JM105 plasmids containing human L1^RP and a mutant in its reverse transcriptase domain, respectively (generous gifts from Dr. John Moran, University of Michigan, Ann Arbor, MI), using Fugene 6 as described by manufacturer (Roche Diagnostics, Indianapolis, IN). Stable transfectants were selected over 2 weeks in the presence of 200 μg/mL hygromycin (Sigma, St. Louis, MO). Individual transfectants were collected, mixed, expanded, and kept frozen at −80°C before individual experiments.

Retrotransposition assay. The approach employed was first applied by Curcio and Garfinkel (30) to examine Ty1 retrotransposon insertions in yeast and later adapted for study of the L1 retrotransposition in mammalian cells (10). In this assay, human L1^RP is tagged with a reporter cassette containing an antisense copy of the antibiotic resistance gene (neo) and interrupted by an intron in the sense orientation. Transcription from either the L1 5′-UTR or a heterologous promoter (P1), splicing of the intron, reverse transcription, and insertion of the cDNA into chromatin allow stable expression of the reporter gene and acquisition of geneticin (G418) resistance. The specificity of the response can be monitored based on the inability of transcripts originating from the P2 promoter to splice intronic sequences and synthesize neogene product. Consequently, cells remain sensitive to G418 and fail to form colonies when grown in culture over sustained periods. Cells (∼1 × 10⁶) were seeded on 10-cm plates, allowed to attach for 24 hours, and treated with either 3 μmol/L BaP or DMSO vehicle (Sigma) for up to 72 hours. Cultures were trypsinized and counted, and the same number of cells (5 × 10⁵ per 10-cm plate) was reseeded on new plates. After recovery for 24 hours, G418 selection (400 μg/mL; Invitrogen) was initiated for 14 days with medium change every third day. Visible G418-resistant foci were fixed and stained as described (31). For measurements of cellular viability, triplicate cultures were trypsinized at days 3 and 6 of G418 selection and live cells were counted after staining with 0.4% trypan blue (Sigma).

RNA analysis. Total RNA was isolated from cells treated with BaP or DMSO for 24, 48, or 72 hours using Trizol (Invitrogen). RNA samples were treated with DNase I (DNA-free kit, Ambion, Austin, TX). RNA (0.5 μg) was reverse transcribed in 20 μL with SuperScript II (Invitrogen) using oligo(dT). The reverse transcription reaction (4 μL) was used for PCR. All primers were annealed at 63°C (30 seconds), extended at 68°C (1 kb/min), and subjected to 40 cycles following denaturation at 94°C (30 seconds). Primer sequences are available upon request.

DNA analysis. Total DNA was isolated from cells treated with BaP or DMSO for 24, 48, or 72 hours using DNazol (Invitrogen). DNA samples were digested overnight with SwaI (New England Biolabs, Beverly, MA), purified using Qiaquick PCR purification kit (Qiagen Sciences, Valencia, CA), and subjected to PCR using the same pair of primers used for RNA analysis. The primers bind immediately after the stop codon of ORF2 and in the first exon of the neocassette as detailed below.

BaP treatment up-regulates L1 RNA in HeLa cells. Expression of L1 retrotransposon sequences is up-regulated in mouse somatic cells challenged acutely with BaP or its oxidative metabolites as determined by Northern blot analysis, differential display, and real-time PCR (27–29). To determine if BaP treatment activates human L1 expression and increases L1 retrotransposition, a plasmid-based strategy was employed using cultured HeLa cells. HeLa stable transfectants harboring human L1^RP under control of a heterologous cytomegalovirus promoter were treated with BaP (3 μmol/L), and total RNA was analyzed by reverse transcription-PCR (RT-PCR). Figure 1 shows an increase in L1 RNA levels on BaP challenge of the cells. Expression of Cyp1a1, a BaP-inducible gene, was up-regulated, indicating that cells responded to xenobiotic treatment. In contrast, expression of c-myc did not change. Therefore, BaP exposure is associated with up-regulation of L1 RNA.

L1^RP cDNA synthesis is up-regulated by BaP. L1 retrotransposition requires transcription from the promoter, intron splicing, reverse transcription, and cDNA insertion into chromatin after second-strand DNA synthesis. Therefore, L1 cDNA synthesis was measured as an index of the final and, most likely, rate-limiting step before insertion. PCR amplification of total DNA isolated from BaP-treated and control cells was employed. Before PCR, total DNA was digested with SwaI. The SwaI restriction site is unique in the region of parental L1 to be amplified and cuts in the β-globin intronic sequence, thus allowing amplification only from L1 cDNA either single or double stranded. PCR amplification was then carried out using oligonucleotides that prime immediately after the stop codon of ORF2 and in the first exon of the neocassette (Fig. 2A). Figure 2B shows that a product of the expected size was detected after 48 and 72 hours of BaP treatment. No changes were observed after 24 hours. The PCR product was not detected in samples from DMSO-treated cells, indicating that increase in L1 was specific for BaP treatment. Sequence analysis of the PCR product confirmed that it was spliced out neo-DNA, thus ratifying the conclusion that BaP up-regulates L1 cDNA synthesis.

BaP treatment sensitizes HeLa cells to G418. We next tested whether increases in L1 RNA and cDNA lead to increased retrotransposition. Three days after BaP treatment, cultures were trypsinized, cells were counted, and the same number of cells was reseeded on new plates. After recovery for 24 hours, G418 selection was initiated. The aminoglycoside antibiotic, G418, is a translational inhibitor that alters ribosomal proofreading and induces cell death. Indeed, vehicle-treated cells undergo ∼2.5 cell divisions in the presence of G418, and cell numbers start to decline some time after day 3 with an average of 9 days required for nonresistant HeLa cells to clear the plate (data not shown). BaP-treated cells were extremely sensitive to G418 treatment as evidenced by increased rates of cell death compared with control (Fig. 3A). The interaction between BaP with G418 may be related to genotoxicity because the halogenated aromatic hydrocarbon 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD), a nongenotoxic aromatic hydrocarbon carcinogen, did not sensitize cells to G418 (data not shown).

BaP increases retrotransposition frequencies. HeLa cell plates were incubated for 2 weeks after BaP or DMSO treatment, and visible foci were fixed and stained with trypan blue. Figure 3B shows that despite the high rates of cell death during G418 selection BaP-treated cells yielded resistant foci at frequencies comparable with control. Thus, when adjusted for cell death rates, the frequency of retrotransposition in BaP-treated cells was significantly higher than in control. To directly compare retrotransposition frequencies, cells from BaP-treated and DMSO-treated cultures were isolated at day 6 when selection-associated cell death was stabilized (Fig. 3C) and were reseeded in new plates for continued G418 selection. Increased retrotransposition frequencies were consistently observed in BaP-treated cultures, indicating that BaP markedly activates L1 retrotransposition in HeLa cells.

To determine whether growth resistance to G418 was mediated by creation of a functional neocassette (which can only be made after L1 transcription, splicing, reverse transcription, and genomic integration), genomic DNA was isolated from expanded resistant colonies and subjected to PCR (see Fig. 3). A 2-kb product from the parental plasmid DNA was observed. All G418-resistant clones yielded a 1-kb product, which can only be obtained after retrotransposition (Fig. 4, lanes 2-7). Neither product could be detected in nontransfected HeLa cells. Sequence analysis identified it as a neospliced gene.

The expression of L1 retrotransposon in vivo is limited almost exclusively to germ-line and embryonic cells (15, 17, 18). L1 retrotransposition has been shown to occur in mouse early embryos (32). Induction of L1 can be observed in embryonic carcinoma cells, testicular germ-line tumors and ovarian carcinomas (33–35), and, to a lesser degree, in other tumors (36).

The cellular mechanisms that mediate L1 activation in somatic cells remain largely obscure. Earlier studies revealed that L1 elements are induced in cultured rat chloroleukemia cells on exposure to UV light and ionizing radiation (20). Radiation-induced activation of L1 may be associated with direct DNA damage. Evidence is presented here that BaP, a genotoxic polycyclic aromatic hydrocarbon carcinogen, increases L1 cDNA synthesis and retrotransposition in HeLa cells. The modulation of mammalian genes by BaP often involves genotoxic and mutational mechanisms dependent on the induction of double-strand DNA breaks caused by reactive intermediates (reviewed in ref. 37). BaP is metabolized in mammalian cells by cytochrome P450s to diol-epoxides and quinones that form covalent DNA adducts and induce oxidative DNA lesions (38). These modifications alter gene expression and activate DNA damage checkpoints associated with G₁ cell cycle arrest and DNA repair (39). BaP modulates gene expression via transcriptional mechanisms that involve aryl hydrocarbon receptor (AhR; ref. 40), activator protein-1 (41), and nrf2 (42, 43) as well as posttranscriptional mechanisms involving gene splicing deficits (44) and mRNA stability (45). Thus, one or more of these mechanisms may contribute to the up-regulation of L1 by BaP in HeLa cells.

L1 sequences often carry substantial DNA methylation and are generally heterochromatinized. Unlike normal cells, HeLa cells may have a generally undermethylated genomic DNA that helps create a local chromatin environment permissive for L1 activation. Under the conditions examined here, however, any contribution of DNA undermethylation to the observed patterns of L1 activation would likely be minimal because significant activation is seen in different cellular contexts as well as in nontransformed lines of endothelial and vascular smooth muscle cell origin (Supplementary Fig. S1). Cellular specificity is observed because L1 activation following carcinogen exposure is absent in the RAW264.7 macrophage-like cell line (Supplementary Fig. S1). The possibility remains that changes in L1 methylation status by stressful stimuli contribute to L1 regulation in mammalian cells.

Previously, we reported that expression of L1 retrotransposon sequences is dramatically up-regulated in mouse vascular smooth muscle cells challenged with BaP or its oxidative metabolites as determined by Northern blot analysis, differential display, and DNA microarrays (27–29). Moreover, a novel retrotransposon of mouse L1Md-A2 lineage was found in vascular smooth muscle cells that contain two antioxidant responsive elements in its 5′-UTR region and participate in transactivation (29). Insertion of this 5′-UTR upstream of a reporter gene increases reporter gene expression 2-fold in response to the genotoxic carcinogen BaP and the nongenotoxic carcinogen TCDD. Interestingly, the human L1^RP 5′-UTR sequence studied here also contains putative binding sites for proteins that bind and activate the antioxidant (5′-TGACGGACG-3′) and the AhR response elements (5′-GCGTG-3′) located at positions +130 to +138 and +54 to +58 relative to the transcriptional initiation start site of the gene, and induction of exogenous and endogenous L1 is observed in HeLa cells.¹

Thus, redox-dependent activation of L1 occurs in mammalian cells in vitro and may mediate hydrocarbon inducibility. Although both BaP and TCDD activate L1 promoter luciferase constructs in transient transfection experiments, only BaP enhances L1 retrotransposition in HeLa cells. Thus, DNA damage by BaP may play a role in L1 retrotransposition. Indeed, DNA-damaging agents, such as cisplatin, etoposide, and γ-irradiation, substantially induce endogenous reverse transcriptase activity in murine cells (21). This increase in endogenous reverse transcriptase activity comes from L1 because it mediates retrotransposition of exogenous AluI element (25). In addition to compounds that damage DNA directly, several heavy metals, considered to be carcinogenic, also stimulate L1 retrotransposition in HeLa cells (46).

Recent evidence has established a direct association between L1 retrotransposition and genomic instability as determined by element inversions, extranucleotide insertions, exon deletions and exon shuffling, chromosomal inversions, and flanking sequence comobilization (23, 47–49). The induction of L1 elements on stressful challenge may represent a misguided attempt at DNA repair. This interpretation is consistent with studies showing that retrotransposon-derived reverse transcriptase can mediate repair of double-strand chromosomal breaks (50). Furthermore, double-strand breaks in chromosomal L1 elements can be repaired by homologous recombination with nonallelic endogenous elements, leading in some instances to gene conversion between repetitive elements (51). Increased prevalence of L1 retrotransposon sequences at sites of DNA double-strand break repair in mouse models of metabolic oxidative stress and MYC-induced lymphoma suggests a role for L1 in DNA rearrangements (52). In addition, involvement of L1-encoded products in AluI transposition has been recently shown (24, 25), suggesting that activation of L1 by genotoxic mutagens may also contribute to chromosomal rearrangements by transposing AluI sequences. Genetic events involving recombination of AluI repeats have been implicated in human genetic diseases (53).

The activation and reintegration of retrotransposons into the genome has been linked to several diseases in human and rodents; however, mechanisms remain largely unknown (54). A total of 48 retrotransposon or retrotransposon-mediated insertions (15 for L1, 25 for AluI, and 4 for SVA elements) have been identified in connection to human disease (55). Although AluI sequences are transposed by active L1 elements (24–26), experimental evidence is lacking for SVA sequences. Thus, activation of retrotransposition frequencies in BaP-treated somatic cells suggests that epigenetic dysregulation of retroelements may contribute to carcinogen-induced mutations, genomic rearrangements, and genomic instability. The potential role of L1 activation in human disorders, such as cancer and atherosclerosis, in which DNA damage coupled to deficits in DNA repair and genomic instability mediates shifts in phenotypic stability of target cells, needs to be investigated.

Endogenous reverse transcriptase is involved in proliferation and differentiation of tumorigenic and nondifferentiated cells (56) and is essential in murine early embryo development (57). Inhibition of reverse transcriptase causes developmental arrest during preimplantation, which is associated with reprogramming of gene expression (57). Moreover, down-regulation of L1 reverse transcriptase expression with small interfering RNA induces a differentiated morphology in transformed cells (58), and pharmacologic inhibition of reverse transcriptase reduces the growth of human tumor xenografts in athymic nude mice (58). Based on these results, it has been proposed that endogenous reverse transcriptase modulates expression of genes that promote the transition from low proliferating, differentiated phenotypes to highly proliferative, transformed phenotypes.

Evidence is presented here that BaP increases L1 retrotransposition in HeLa cells following up-regulation of L1 RNA levels, increased L1 cDNA synthesis, and stable genomic integration. These findings suggest that the mutations and genomic instability observed in human populations exposed to genotoxic environmental hydrocarbons may involve epigenetic activation of mobile elements dispersed throughout the human genome.

Grant support: National Institute of Environmental Health Sciences grant R01 ES04849-15.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.