Undifferentiated cells and embryos express high levels of endogenous non-telomerase reverse transcriptase (RT) of retroposon/retroviral origin. We previously found that RT inhibitors modulate cell growth and differentiation in several cell lines. We have now sought to establish whether high levels of RT activity are directly linked to cell transformation. To address this possibility, we have employed two different approaches to inhibit RT activity in melanoma and prostate carcinoma cell lines: pharmacological inhibition by two characterized RT inhibitors, nevirapine and efavirenz, and downregulation of expression of RT-encoding LINE-1 elements by RNA interference (RNAi). Both treatments reduced proliferation, induced morphological differentiation and reprogrammed gene expression. These features are reversible upon discontinuation of the anti-RT treatment, suggesting that RT contributes to an epigenetic level of control. Most importantly, inhibition of RT activity in vivo antagonized tumor growth in animal experiments. Moreover, pretreatment with RT inhibitors attenuated the tumorigenic phenotype of prostate carcinoma cells inoculated in nude mice. Based on these data, the endogenous RT can be regarded as an epigenetic regulator of cell differentiation and proliferation and may represent a novel target in cancer therapy.
A striking finding emerging from the recent sequencing of the human genome is that retrotransposable elements, such as long interspersed elements (LINEs), Alu and endogenous retroviruses (ERVs) make up some 45% of human DNA (Deininger et al., 2003). All classes of retroelements, but the Alu family, are endowed with a reverse transcriptase (RT)-coding gene, which enables them to retrotranspose autonomously. The presence and function of retroelements in the genome has long puzzled biologists. The lack of any obvious cellular function initially suggested that these elements, and the RT-coding sequences harbored therein, were mere evolutionary remnants and inspired the concept of ‘junk DNA’. More recently, the realization that retrotransposons can reshape the genome and contribute to modulation of gene expression has led to reconsider that hypothesis. Growing evidence indicate that RT-coding genes are expressed at low levels, if at all, in differentiated nonpathological tissues; in contrast, high expression is distinctive of germ cells (Kiessling et al., 1989; Giordano et al., 2000), embryos (Poznanski and Calarco, 1991; Packer et al., 1993), embryonic tissues (Mwenda, 1993), and undifferentiated and transformed cells (Deragon et al., 1990; Martin, 1991; Martin and Branciforte, 1993), suggesting that levels of RT expression are linked to the proliferative potential of the cell. In addition, RT gene activity is upregulated by a variety of stimuli acting at the genome-wide level, for example, cellular stress (Hagan et al., 2003), heat shock, cycloheximide, adenovirus infection (Li and Schmid, 2001), genotoxic agents (Hagan and Rudin, 2002), and DNA base analogs (Khan et al., 2001). Unscheduled activity of retrotransposons and ERVs is implicated in a variety of diseases, including cancer (Friedlander and Patarca, 1999; Ostertag and Kazazian, 2001). Conversely, inactivation of specific RT-encoding elements using antisense oliogonucleotides or ribozymes inhibited proliferation of human (Kuo et al., 1998) and murine (Crone et al., 1999) cell lines. The question remains unanswered as to whether retroelements are to be regarded as endmarkers, or causative triggers, in processes associated with cell proliferation and function.
We showed previously that non-nucleosidic RT inhibitors directed against the HIV-encoded RT in fact also inhibit the endogenous RT activity present in early embryos (Pittoggi et al., 2003) and undifferentiated cells (Mangiacasale et al., 2003). These RT inhibitors displayed powerful effects in nondifferentiated, or dedifferentiated cells: they inhibited cleavage in murine early embryos, yielding developmental arrest before the blastocyst stage (Pittoggi et al., 2003), and, moreover, reduced proliferation and facilitated the onset of differentiation in murine and human cell lines (Mangiacasale et al., 2003). These data suggest that the endogenous cellular RT plays a physiological role in cell proliferation and differentiation. Furthermore, there appears to be a requirement for regulated RT activity in differentiated cells. Here we have taken a step further and have asked whether the endogenous RT plays a direct role in proliferation and differentiation of transformed cells, and, if so, whether modulating RT activity in cancer cells might represent a novel approach to inhibit tumor growth.
RT inhibitors reversibly reduce cell proliferation in human transformed cell lines
In previous work, we reported that the RT inhibitor nevirapine, widely used in anti-HIV therapy, blocks the enzymatic activity of endogenous RTs in noninfected proliferating cells, as revealed using a highly sensitive RT–PCR based in vitro assay (Pyra et al., 1994), and concomitantly reduces the rate of proliferation (Mangiacasale et al., 2003). Here we set out to investigate the response of human transformed cell lines to prolonged exposure to RT inhibitors. Two well-characterized RT inhibitors, that is, nevirapine and efavirenz, were used. Cells from A-375 melanoma, PC3 prostate carcinoma and TVM-A12 primary melanoma-derived lines were passaged, counted and replated every 96 h with continuous drug readdition (or DMSO alone in control cultures) for at least five 96 h-cycles. As shown in Figure 1a, both inhibitors effectively reduce cell growth in all cell lines, with a stable effect during prolonged exposure. Growth inhibition was reversible: when RT inhibitors were removed, proliferation was resumed at a comparable rate to controls within one or two 96 h-cycles. Readdition of the drugs inhibited again proliferation in all cell lines. Thus, the reduction of cell growth associated with RT inhibition is not inherited as a permanent change through cell division.
We next asked whether either RT inhibitor induced cell death in A-375 or PC3 cell lines. Combined staining with PI to reveal permeable necrotic cells, DAPI to visualize apoptotic nuclei, and DiOC6(3) to monitor the loss of mitochondrial transmembrane potential, revealed no significant induction of cell death by either RT inhibitor; what low ratio was recorded (15% at most after 72 h of exposure to either drug) was largely accounted for by apoptosis (data not shown). Thus, neither drug has significant nonspecific toxicity. We next sought to establish whether reduced cell growth rather reflected the induction of cell cycle delay. Biparametric FACS analysis was employed to measure the DNA content (revealed by PI) and DNA replication (by BrdU incorporation) after four 96 h-cycles of exposure to RT inhibitors. This depicted significantly altered cell cycle profile in anti-RT treated cultures, with an increased proportion of G0/G1 BrdU-negative cells, that was especially pronounced in A-375 cell cultures (Figure 1b). Removal of the drugs re-established the original cell cycle profile and abolished the G1 delay.
Nevirapine induces morphological differentiation and modulates gene expression in transformed cell lines
Melanomas are resistant to most therapeutic treatments: thus, it was relevant to determine whether RT inhibitors induced differentiation concomitant with reduced cell growth. We first examined A-375 melanoma cells, which acquire a typical dendritic-like phenotype in response to certain inducers of differentiation (Sauane et al., 2003). As shown in Figure 2A, morphological differentiation, revealed by cell shape, dendritic-like extensions and increased adhesion, became evident within 4–5 days of exposure to nevirapine (d) or efavirenz (g), compared to DMSO-treated controls (a). By scanning electron microscopy (SEM), A-375 cells cultured with nevirapine (e) and efavirenz (h) become flattened compared to untreated controls (b) and exhibit elongated dendrite extensions that adhere tightly to the substrate. Confocal microscopy after α-tubulin immunofluorescence also revealed the reorganization of microtubule arrays throughout the length of outgrowing dendrites in RT-inhibited cells (f, i), different from controls (c), in which short microtubules concentrate around the nucleating centers. Nevirapine treatment induced similar changes in TVM-A12 primary cells derived from melanoma (Figure 2B): untreated cells have a spindle-shaped morphology by phase contrast (a) and SEM (b); nevirapine-treated TVM-A12 cells formed instead typical branched dendrites (d, e) and displayed well-organized, elongated microtubule arrays (f), compared to untreated cells (c). Significant morphological changes were also induced in PC3 prostate carcinoma cells upon exposure to nevirapine (Figure 3c and d) and efavirenz (e, f) compared to controls (a, b). The microtubule network was reorganized, with the appearance of fusiform extensions protruding from the cell periphery, particularly in response to nevirapine.
The induction of morphological differentiation suggests that critical regulatory genes are modulated in response to the RT inhibitory treatment. This was investigated in semiquantitative RT–PCR analysis of cultures treated with DMSO only, or nevirapine or efavirenz for four cycles. In A-375 melanoma cells, we focussed on a set of four genes: the E-cadherin gene, involved in cell–cell adhesion and expressed in differentiated but not in tumor cells (Hsu et al., 2000); and the c-myc, bcl-2 (Utikal et al., 2002) and cyclin D1 (Sauter et al., 2002) genes, which are directly implicated in melanoma cell proliferation and tumor growth. Results in Figure 4a indicate that the E-cadherin gene is markedly upregulated, whereas c-myc, bcl-2 and ccnd1 genes are downregulated, in RT-inhibited A-375 cultures compared to controls. One exception was recorded for efavirenz, which failed to downregulate ccnd1 expression. We also analysed PC3 prostate carcinoma cells and selected two marker genes of differentiated prostate epithelia, that is, the prostate-specific antigen PS-A (Lilja, 2003) and androgen receptor (AR) (Linja et al., 2001) genes. Neither of these genes is expressed in untreated cultures, yet both genes were induced in response to RT inhibitors (Figure 4b). Again, the expression of all genes returned to the original level when the inhibitors were removed. Thus, RT inhibitory drugs modulate the expression of critical genes in transformed cells, consistent with the induction of differentiation, yet this reprogramming is reversible and is abolished when RT-inhibition is released.
RNA interference (RNAi) against RT-encoding LINE-1 elements reduces proliferation and promotes differentiation in melanoma cells
At this point, we wanted to ascertain unambiguously whether reduction of cell growth and induction of differentiation by pharmaceutical RT inhibitors are specifically attributable to the inhibition of the cellular RT. To address this question, RNAi experiments were designed to target specifically LINE-1 element families that are most abundantly expressed in human cells (Brouha et al., 2003, and Supplementary information). Double-stranded RNA oligonucleotides directed against LINE-1 ORF1 (L1-i, Figure 5a), or the lamin A/C gene (lam), or representing a noninterfering sequence (n.i.), were transfected in A-375 cells. Semiquantitative RT–PCR analysis 24, 48 and 72 h after transfection indicated that LINE-1 expression is progressively reduced in LINE-1-interfered cultures (Figure 5b, lanes 2, 4 and 6) compared to controls treated with noninterfering (lanes 1, 3 and 5), or lamin RNAi (lane 7) in which lamin expression is abolished (lane 10). Conversely, lamin gene expression was not affected in LINE-1-interfered cells (lane 9). Neither LINE-1 nor lamin RNAi influenced GAPDH expression. Thus, our RNAi conditions target specifically LINE-1 expression. At 72 h after L1-i transfection, expression of both ORF1 and ORF2 was reduced by almost 80% compared to cells transfected with noninterfering oligonucleotide (Figure 5c). To ascertain that the RT protein product was consistently downregulated in L1-i interfered cells, we made use of a recently developed antibody against LINE-1 ORF2-encoded RT (termed EN-L1, Ergun et al., 2004). Immunoprecipitation and Western blot assays indicate that RT protein levels 72 h after transfection are significantly reduced in L1-i cells (Figure 5d). Remarkably, L1-i interfered cultures, in which the RT protein was downregulated, developed a typical differentiated morphology (Figure 5e), concomitant with reduced cell growth (Figure 5f), compared to cells transfected with noninterfering oligonucleotide. Furthermore, LINE-1 interference induced downregulation of expression of c-myc and cyclin D1 genes, but not of GAPDH (Figure 5c). These effects are comparable to those obtained with pharmacological RT inhibition.
RT inhibitors reduce the growth of human tumor xenografts in athymic nude mice
Given that proliferation and differentiation can be modulated by RT inhibition in transformed cells, we next asked whether RT inhibitors also affect tumor growth in vivo. Tumorigenic cell lines selected for these experiments include A-375 and PC3 lines, as well as HT29 colon and H69 small cell lung carcinoma lines, which also showed reduced cell growth in response to RT inhibitors (Mangiacasale et al., 2003, and data not shown). Cells were inoculated subcutaneously in the limb of athymic nude mice. Animals were then subjected to treatment with efavirenz, because this drug showed a higher in vivo effectiveness than nevirapine in preliminary assays. The optimal dose (20 mg/kg body weight) was determined in dose-response experiments testing 4–40 mg/kg of the drug. The efavirenz treatment proved safe in all animal groups, with no animal death or explicit signs of toxicity in any of the groups – although the group treated with 40 mg/kg showed a significant decrease of body weight in more than 60% of animals. Figure 6 shows the recorded curves of tumor growth in mice untreated (red) or treated with efavirenz, starting one day (purple), or 1 week (yellow), after tumor inoculation. Tumor growth was markedly reduced in treated compared to untreated animals for all xenograft types, and tumor progression was antagonized with comparable effectiveness regardless of the timing of the treatment start, despite of differences in the initial tumor size. The growth curves of PC3- and HT29-derived tumors in animals treated from day one after inoculation, but subjected to treatment discontinuation after day 15 (green curves), indicate that the inhibition of tumor growth requires continuous RT inhibition in vivo.
Efavirenz-treated PC3 cells exhibit reduced tumorigenicity in vivo
Finally, we asked whether pretreatment of transformed cells with efavirenz modifies the tumorigenic potential of derived xenografts. PC3 prostate cancer cells were cultured with efavirenz for two 96 h cycles, a time that was sufficient for induction of the PS-A and AR genes in cultured cells (Figure 4b), and subsequently inoculated in nude mice. Untreated cells were inoculated in parallel batches of animals. Efavirenz-pretreated, or untreated, PC3 cell xenografts were then either continuously treated with efavirenz in vivo or were left untreated. Figure 7a shows the rate of tumor growth in these experiments: untreated PC3 cells developed fast-growing tumors in all animals. In contrast, efavirenz-pretreated PC3 cells showed a significantly reduced tumor-forming ability in vivo, and xenografts grew more slowly. Figure 7b shows the incidence of tumors in animals models exposed to different treatments: while 100% of inoculated animals developed aggressive tumors using untreated PC3 cells, efavirenz-pretreated cells developed slowly-growing xenografts in 65% of the inoculated animals. Moreover, only 40% of the animals inoculated with pretreated cells and further treated with efavirenz in vivo developed a tumor at all, and in that case the growth curve was flat. Thus, pretreatment of cells with anti-RT drugs before inoculation attenuates the tumorigenic potential of transformed cells.
This work highlights three unexpected aspects of the human genome that have implications for cancer: first, LINE-1 elements are identified as active components of a mechanism involved in control of cell differentiation and proliferation; second, RNAi-dependent inactivation of LINE-1 elements, or pharmacological inhibition of the endogenous RT activity which they encode, can restore control of these traits in transformed cells; third, inhibitors of RT reduce tumor growth in animal models in vivo.
The RT inhibitors used here, nevirapine and efavirenz, share a common biochemical mechanism of action by binding the hydrophobic pocket in the p66 subunit of RT enzymes (Di Marzo Veronese et al., 1986; Ren et al., 2001). Though being designed to target the HIV-encoded RT, nevirapine proved able to inhibit the endogenous retrotranscriptase activity present in noninfected cells (Mangiacasale et al., 2003) in a highly sensitive in vitro assay (Pyra et al., 1994). We now show that both drugs reduce proliferation of transformed cells, largely independent of cell death, but associated with G1 delay or arrest. Concomitant with this, RT inhibitors induce morphological differentiation of transformed cells. The induction of differentiation is rapid, different from phenotypic changes elicited by inhibitors of the telomerase-associated RT (TERT), which require long treatment times (120 days) (Damm et al., 2001). Furthermore, we never observed the reorganization of actin stress fibers or focal adhesion sites typical of senescent cells. The absence of senescence-specific modifications, and the rapid induction of differentiation, indicate that the RT inhibitors do not target TERT and induce a low-proliferating differentiated phenotype rather than senescence.
That these effects are specifically associated with RT inhibition was further demonstrated in RNAi experiments targeting active LINE-1 retroposon families accounting for 84% of the overall retrotransposition capability in human cells (Brouha et al., 2003). In A-375 cells, our RNAi conditions specifically downregulate expression of LINE-1 ORF1 and ORF2 by some 80%, suggesting that the biologically active LINE-1 subgroup was effectively silenced. Consistently, RT protein levels were also downregulated in LINE-1-interfered cells. RNAi to RT encoding LINE-1 elements induced morphological, proliferative and gene-profiling changes that are virtually indistinguishable from those caused by pharmacological RT inhibitors. The similarity of phenotypes induced by independent approaches indicates that inhibition of LINE-1 expression, or of RT activity, is sufficient to delay proliferation and promote differentiation. These observations rule out that nonspecific side effects of the drugs cause the observed cell phenotypes and highlight the specificity of the role of RT.
Consistent with growth reduction and induction of differentiation, RT inhibition caused the reprogramming of gene expression: this implicates the endogenous RT in modulation of expression of genes that promote the transition from highly proliferating, transformed phenotypes to low proliferating, differentiated phenotypes, suggesting that genome function is the ultimate target of pharmaceutical or RNAi-dependent inhibition of RT activity. Changes in gene expression are not inherited through cell division, but are reversible when RT inhibition is released. The reversibility of examined features after release of the inhibition suggest that LINE-1 encoded RT is part of an epigenetic mechanism that can modulate gene expression and contributes to the molecular mechanisms underlying cell proliferation and differentiation.
A relevant finding in this study is the ability of RT inhibitors to reduce tumor growth in nude mice inoculated with four human xenograft models in vivo. Tumor growth was inhibited as long as the animals were supplied with RT inhibitor, yet was resumed on discontinuation of the treatment, as observed in cell lines, consistent with an epigenetic role of the endogenous RT in tumor growth. These data illustrate the promising cytostatic ability of RT inhibitors in cancer treatment. Furthermore, in vitro pretreatment of PC3 prostate carcinoma cells with efavirenz attenuates their tumorigenicity in vivo. Thus, the activation of differentiation markers and reduced proliferation associated with RT inhibition are part of a large-scale reprogramming that can attenuate the malignant phenotype of transformed cells in vivo.
Growing data indicate that epigenetic changes can reprogram tumor cells and convert the transformed phenotype into a ‘normal’ non pathological state (Lotem and Sachs, 2002; Li et al., 2003). Epigenetic reprogramming can bypass the genetic alterations that originally caused the malignant transformation in a variety of tumors (Lotem and Sachs, 2002). Therefore, epigenetic regulatory factors are viewed as valuable, worth-challenging targets in tumor therapy (Egger et al., 2004). Retrotransposons can contribute to heterochromatin formation in fission yeast (Schramke and Allshire, 2003). Though such a mechanism has not yet been proved in higher eukaryotes, unpublished results in our laboratory suggest that LINE-1-encoded RT is implicated in nuclear reorganization and positioning of specific genes. Importantly in this respect, the intranuclear position of genes is directly linked to their expression (Osborne et al., 2004 and references herein).
Of relevance to the present study is the observation that – while many tested compounds targeting the ‘epigenome’ have generally proven toxic and/or chemically unstable – both nevirapine and efavirenz have been used in AIDS treatment for many years. Thus, the prospect of using these RT inhibitors in cancer therapy would have obvious advantages given their epidemiological record of generally good tolerance to continued administration. In retrospect, epidemiological evidence indicate that Kaposi's sarcoma (Portsmouth et al., 2003) and other AIDS-related cancers (Tirelli and Bernardi, 2001) have a reduced incidence in patients treated with highly active antiretroviral therapy (HAART): while this is generally viewed as a reflection of the improved immune reaction in treated patients, it may also suggest a direct inhibitory effect of HAART on the endogenous RT activity in tumor cells.
Materials and methods
Human A-375 melanoma (ATCC-CRL-1619), TVM-A12 primary melanoma-derived (Melino et al., 1993), HT29 adenocarcinoma (ATCC HTB-38), H69 small-cell-lung carcinoma (SCLC) (ATCC HTB119), and PC3 prostate carcinoma (ATCC CRL-1435) cell lines were seeded in six-well plates at a density of 104 to 5 × 104 cells/well and cultured in DMEM or RPMI 1640 medium with 10% fetal bovine serum. Nevirapine and efavirenz were purified from commercially available Viramune (Boehringer-Ingelheim) and Sustiva (Bristol-Myers Squibb) as described (Pittoggi et al., 2003). The drugs were made 350 and 15 μM in dimethyl sulfoxide (DMSO, Sigma-Aldrich), respectively, and added to cells 5 h after seeding; the same DMSO volume (0.2% final concentration) was added to controls. Fresh RT inhibitor-containing medium was changed every 48 h. Cells were harvested every 96 h, counted in a Burker chamber (two countings/sample) and replated at the same density.
Cell cycle and cell death analysis
BrdU (20 μM) was added to the cultures during the last 30 min before harvesting. Harvested cells were then treated with anti-BrdU antibody and propidium iodide (PI) and subjected to biparametric analysis of the DNA content and BrdU incorporation in a FACStar Plus flow-cytometer (Beckton-Dickinson). Cell death was assessed by microscopy after combined staining with DAPI (nuclear morphology); PI (cell permeability); and 3,3 dihexyl-oxacarbocyanine [DiOC6(3)], a fluorescent probe for mitochondrial transmembrane potential.
Indirect immunofluorescence and confocal microscopy
Cell preparations were fixed with 4% para-formaldehyde for 10 min and permeabilized in 0.2% Triton-X100 in PBS for 5 min. Mouse monoclonal anti-bovine α-tubulin (Molecular Probes, A-11126) was revealed by Alexa Fluor 488-conjugated secondary antibody (Molecular Probes, A-11001) in A-375 and TVM-A12 cells and FITC-conjugated secondary antibody (Jackson Immunoresearch, cat 115-095-068) in PC3 cells. Nuclei were stained either with 2 μg/ml PI in the presence of 0.1 mg/ml ribonuclease A or with 0.1 μg/ml DAPI. Samples were imaged under a confocal Leica TCS 4D microscope equipped with an argon/kripton laser. Confocal sections were taken at 0.5–1 μm intervals.
Scanning electron microscopy (SEM)
Cells were fixed in 2.5%. glutharaldehyde in 0.1 M Millonig's phosphate buffer (MPB). After washing, cells were postfixed with 1% OsO4 (1 h, 4°C) in MPB and dehydrated using increasing acetone concentrations. Samples were critical-point dried using liquid CO2 and sputter-coated with gold before examination on a Stereoscan 240 scanning electron microscope (Cambridge Instr., Cambridge, UK).
RNA extraction and treatment with RNase-free DNase I were as described (Pittoggi et al., 2003). cDNAs were synthesized using 300 ng of RNA, oligo (dT) and the Thermoscript system (Invitrogen). Reaction mixtures (1/25) were amplified using the Platinum Taq DNA Polymerase kit (Invitrogen) and 30 pmol of oligonucleotides (MWGBiotech, Ebersberg, Germany; see Supplementary information) in an initial 2-min step at 94°C, followed by cycles of 30 s at 94°C, 30 s at 58–62°C, 1 min at 72°C. Each oligo pair was used in sequential amplification series with increasing numbers (25–40) of cycles. PCR products were electrophoresed, transferred to membranes and hybridized for 16 h at 42°C with [32P]γ-ATP end-labeled internal oligonucleotides. The intensity of the amplification signal was measured by densitometry in at least three independent experiments for each gene and normalized to the GAPDH signal in the same experiment.
Four 21-nt double-stranded siRNA oligonucleotides encompassing region 1367–2056 in LINE-1 were designed to target the consensus sequence of active LINE-1 elements described by Brouha et al. (2003). The siRNA oligonucleotide targeting bases 2035–2056 (L1-i) was most effective and was used in all experiments. Control cells were treated with noninterfering oligonucleotide (n.i.), 3′-fluorescein-conjugated to monitor transfection efficiency, or with specific siRNA against the lamin A/C gene. All siRNAs were synthesized by Qiagen USA. A-375 melanoma cells were transfected using RNAiFect transfection reagent (Qiagen) and 300 nM of siRNA. Details for siRNA assays are in Supplementary information.
Western blot and immunoprecipitation
At 72 h after siRNA transfection A-374 cells were harvested in PBS with 0.1 mM. PMSF and lysed in lysis buffer (50 mM Tris-HCl, pH 8.1, 10 mM EDTA, 1% SDS) supplemented with protease inhibitors (1 μM pepstatin, 1 μM leupeptin, 0.1 mM PMSF). After centrifugation at 12 000 r.p.m., 4°C, 15 min, the protein concentration in the supernatants was determined using a Coomassie colorimetric assay (Pierce). Samples (20 μg) were loaded on NuPAGE Novex 10% Bis-Tris gel (Invitrogen), transferred onto membranes and verified by Western immunoblotting using monoclonal anti-α-tubulin (Sigma, T5168) and HRP-conjugated secondary antibody (BIORAD, 170-6516). In total, 500 μg of protein extract were then precleared using 75 μl of protein A-Agarose-50% Slurry beads (Upstate Biotechnology) for 30 min at 4°C. After centrifugation (12 000 r.p.m., 4°C), supernatants were incubated overnight (4°C, with continuous rotation) with chicken polyclonal anti-EN-L1 (Ergun et al., 2004), kindly given by Gerald Schumann (Paul-Ehrlich-Institut, Langen, Germany). Precleared extracts were then incubated with 60 μl of beads (1 h, 4°C) with rotation. After removal of the supernatants, proteins were eluted from the beads in 1% SDS, 0.1 M NaHCO3 and precipitated with 10 volumes of acetone. Pellets were resuspended in TE buffer and loaded on NuPAGE Novex 10% Bis-Tris gel as above. Western blot analysis was carried out using chicken polyclonal anti-EN-L1 (1 : 40 dilution) and donkey HRP-conjugated anti-chicken IgY (Jackson Immunoresearch Laboratories, 703-035-155).
Tumor xenografts and treatment of animals
Athymic nude mice (5 weeks old) (Harlan, Italy), kept in accordance with the European Union guidelines, were inoculated subcutaneously with A-375 melanoma (4 × 106), H-69 (107), PC3 (2 × 106) and HT-29 (106) cells in 100 μl PBS. Mice were subcutaneously injected daily five days a week with Efavirenz (20 mg/kg) using a 4 mg/ml stock in DMSO freshly diluted 1 : 1 with physiological solution. Controls were injected with 50% DMSO. Treatment started 1 day or 1 week after tumor implant, and, where indicated, was discontinued after 14 days. Tumor growth was monitored every other day by caliper measurements; volumes were calculated using the formula length × width × height × 0.52 (Umekita et al., 1996).
We are indebted with Dr Gerald Schumann for kindly providing the antibody against LINE1-encoded RT. We are also grateful to Dr A Mai for drug purification and to Drs R Mangiacasale and S Rutella for cell cycle analysis. This work was supported by Istituto Superiore di Sanità (Grant C3H3 ‘Role of endogenous Reverse Transcriptase in tumor growth and embryo differentiation’).
About this article
Inhibition of LINE-1 retrotransposon-encoded reverse transcriptase modulates the expression of cell differentiation genes in breast cancer cells
Breast Cancer Research and Treatment (2014)