HC11 cells are immortalized mouse mammary epithelial cells (47) originating from mid-pregnancy BALB/c mice, with stem-like or progenitor cell properties (48). HC11 cells were cultured in RPMI 1640 medium containing 10% FBS, 2 mM L-glutamine, 5 mg/ml insulin, 10 ng/ml EGF (47), 100 units penicillin and 100 µg/ml streptomycin (all Thermo Fisher Scientific, Inc.) at 37°C in an incubator supplemented with 5% CO₂. A total of 2.5×10⁵ HC11 cells were transfected at room temperature with 2.5 µg of our previously constructed pNVL-3*/EGFP-INT DNA (40) using the PolyFect^® transfection reagent (Qiagen, Inc.), and then cultured for 16 h under cell culture conditions. Hygromycin B-resistant cell clones were isolated following application of 100 µg/ml hygromycin B for 18 days. Isolated clone cells were further cultured for 20 days in cell culture treated flasks (Corning, Inc.), and used for subsequent experimentation. Notably this plasmid harbors a recombinant VL30 element tagged with an EGFP gene-based retrotransposition indicator cassette, and the expression of EGFP protein occurs solely after a retrotransposition event. Thus, VL30 retrotransposition-positive cells can be enumerated by flow cytometry (40). Our previously isolated Pcl.10 cell clone derived from NIH3T3 mouse embryo fibroblasts transfected with pNVL-3*/EGFP-INT (40), as aformentioned. Pcl.10 cells were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% FBS, 2 mM glutamine and antibiotics (100 units penicillin and 100 µg/ml streptomycin, and maintained as aforementioned. Serum starvation was performed in serum-free medium for 24 h at 37°C. Recovery of 30% confluency HC11 cells treated with 1- or 2 IU tinzaparin pre-incubation for 3 h, and/or 40 pg/ml glucose oxidase (GO) for 24 h, was performed by washing the cells twice with RPMI 1640 medium at room temperature. Tinzaparin sodium (syringe 10,000 anti-factor Xa IU/ml) was obtained from LEO Pharma A/S, and appropriate dilutions were made with culture medium. VEGF treatment at 3–12 ng/ml was performed in both HC11 and Pcl.10 cells with mouse recombinant VEGF A (ImmunoTools GmbH) 24 h after serum starvation. Tinzaparin and/or VEGF treatment was performed at 37°C.

HC11 cl.11 assay cells cultured at the desired confluence on sterilized glass coverslips were fixed with 3.7% paraformaldehyde for 30 min on ice. Then coverslip-samples mounted onto microscope slides were observed and photographed under normal or UV light (Nikon Eclipse E800 Fluorescent Microscope).

All tested hygromycin B-resistant HC11 clone cells were trypsinized, washed with PBS, centrifuged at 1,600 x g for 3 min at room temperature, and resuspended in PBS. Samples of 15,000 clone or non-transfected cells (as the negative control) were used to determine EGFP-positivity by flow cytometry, as previously described (40). The fluorescence intensity thresholds were evaluated using non-transfected cells, and sample fluorescence ≥99.60% was considered as negative, and 0.4% as false positive. Above the false-positive threshold, samples were considered as EGFP- or VL30 retrotransposition-positive. The continuous cell proliferation rate (cell index) was measured using a microelectronic biosensor system (xCELLigence^® real-time cell analysis DP) in E-plates using 3,500 seeded cells in a volume of 60 µl RPMI (supplemented with 10% FBS) for up to 80 h. Cellular proliferation was assessed by seeding 60,000 cells per well (~10% confluent) into a 6-well plate. Cells from three wells were trypsinized, stained with trypan blue and counted using a Neubauer hemocytometer every 24 h for a period of 5 days (44).

The foci formation assay was based on a previously described protocol (49). Briefly, 6-well plates bearing a 0.66% nobble agar-base layer (prepared in 10% FBS/RPMI medium supplemented with 100 units penicillin and 100 µg/ml streptomycin), a 0.33% upper- and 0.33% feeder layer, were used. The upper layer was prepared after mixing nobble agar with trypsinized cells from either control or test cell groups, in the presence or absence of 2 IU/ml tinzaparin, respectively. Dishes were incubated at 37°C in a 5% CO₂ cell culture incubator, and supplemented with 200 µl medium every two days to avoid desiccation. The examination of dishes for cell foci was formation performed with an optical microscope.

For mammosphere preparation, VL30 retrotransposition- tumorigenic HC11 cl.19 cells cultured in normal culture plates were trypsinized and seeded into non-adherent plates at 30% density. Following 20 days of culture, the cells were replenished with RPMI-1640 medium every 2 days, and the resulting anchorage-independent/floating mammospheres in the culture medium were collected using a pipette for further use.

A total of 1.5×10⁶ HC11 cl.19 cells were trypsinized, washed with PBS and centrifuged at 1,600 x g for 3 min at room temperature. Cell pellets were resuspended in 800 µl PCR wash buffer [10 mM Tris-HCl (pH 8.4), 50 mM KCl, 1.5 mM MgCl₂ and 0.001% (w/v) gelatin], and centrifuged once more. The pellets were then gently resuspended in 100 µl PCR lysis buffer [10 mM Tris-HCl pH 8.4, 50 mM KCl, 1.5 mM MgCl₂, 0.001% gelatin, 0.1% (v/v) Triton X-100, 0.45% (v/v) Nonidet P-40 and 0.45% (v/v) Tween-20], heated at 80°C for 10 min, cooled at room temperature for 15 min and then treated with 4 µl proteinase K (10 mg/ml) overnight at 55°C. Following heat inactivation at 95°C for 15 min, the samples were centrifuged at 1,600 x g for 3 min at room temperature, and the supernatant, representing an extracted DNA lysate, was collected. A volume of 3–5 µl DNA lysate was used for PCR analysis. PCR reactions were performed using 2.5 units Taq polymerase (Invitrogen; Thermo Fisher Scientific, Inc.) in a final volume of 50 µl, with specific EGFP primers: GFP968 forward, 5′-GCACCATCTTCTTCAAGGACGAC-3, and GFP1013 reverse, 5′-TCTTTGCTCAGGGCGGACTG-3′ (50) using conditions previously reported (40).

Quantitation of endogenous reverse transcriptase (enRT) and VL30 transcripts was performed by RT-PCR. Total RNA was extracted from untreated cells, cells treated with VEGF alone or VEGF/tinzaparin using a RNeasy Mini Kit (Qiagen, Inc.). Subsequently, 100 ng cDNA was prepared from 1 µg total RNA. A set of degenerate primers designed to target the enRT conserved domains (4 and 5 amino acid sequences identified in the amino-terminal coding regions of most known enRT polymerases including that of the Moloney murine leukemia virus) (51), as well as a set of previously reported VL30-specific primers: Forward, 5′-CCTTTGTTGCCCAGGTAAGTC-3′ and reverse, 5′-CACTGTAGCCAGTTGTGACCAG-3′ (52). mRNA expression of the mouse β-actin gene was quantitated using the following primers: β-actin forward, 5′-TTGCTGACAGGATGCAGAAG-3′ and reverse, 5′-ACATCTGCTGGAAGGTGGAC-3′. The enRT thermocycling conditions were as follows: 94°C for 4 min; 27 cycles of 94°C for 30 sec, 50°C for 45 sec, and 72°C for 2 min; and 72°C for 2 min. The VL30 thermocycling conditions were: 94°C for 4 min; 30 cycles of 94°C for 30 sec, 57°C for 30 sec and 72°C for 2 min; and 72°C for 2 min. Finally, the β-actin conditions were: 94°C for 4 min; 27 cycles of 94°C for 30 sec, 57°C for 30 sec, and 72°C for 1.5 min; and 72°C for 2 min. All PCR products were fractionated using 1.2% (w/v) agarose gels, stained with ethidium bromide (EtBr), and visualized under ultraviolet light. enRTs and VL30 RNA expression values were normalized to that of β-actin, and densitometric analysis was performed using Fiji (ImageJ2) software (National institutes of Health).

Statistical analysis was performed using GraphPad Prism version 5.0 (GraphPad Software, Inc.). Comparisons between multiple groups of retrotransposition frequency values were determined by one-way ANOVA followed by the Tukey's post hoc-test. A paired Student's t-test was used to assess statistically significant differences between two groups. P<0.05 was considered to indicate a statistically significant difference.

In a previous study, VL30 retrotransposition-positive HC11 cell clones were generated, exhibiting an induced EMT phenotype, CSC properties and solid tumor production in syngeneic Balb/c mice (46), and with a retrotransposition frequency of ~2-5.5%. Notably, as HC11 cells possess stem-like properties, the occurrence of retrotransposition events could be explained by their hypomethylation status (46), as has been suggested for human stem cells (53). To examine the effect of tinzaparin and/or VEGF treatment on VL30 retrotransposition, new clones were isolated with a retrotransposition frequency of 7–15%, allowing a more accurate measurement of retrotransposition changes, especially in the case of inhibition.

Following transfection with pNVL-3*/EGFP-INT, 20 hygromycin B-resistant HC11 clones were isolated. Using non-transfected HC11 cells as the control, the isolated clones were flow cytometrically examined for retrotransposition positivity, as exemplified for clone 11 (Figs. S1A and S2). Among all clones examined, a set of seven clones (cl.) exhibited the following percentages of retrotransposition frequency: cl.12, 7%; cl.15, 8%; cl.6 and cl.9, ~9%; cl.11, 11.5%; cl.5, 12%; and cl.17, 15.2%, falling within the range of 7–15.2% (Fig. S1B). PCR analysis confirmed retrotransposition-positivity at the genomic level in four representative clones, through the diagnostic intron-less 342 bp band [indicative of a retrotransposition event (40) (Fig. S1C)]. Furthermore, microscopic observation revealed two common VL30 retrotransposition-derived features, as shown for representative cl.11 (Fig. S3): i) An expected VL30 retrotransposition-induced EMT phenotype (46) observed at low confluence (Fig. S3B); and ii) cytoplasmic EGFP fluorescence (Fig. S3D), confirming the occurrence of VL30 retrotransposition events (40), associated with genomic instability manifested by multinucleated cells bearing enlarged cytoplasm, in cell clusters formed at high confluence (Fig. S3C). Thus, isolated clones with a retrotransposition value of 7–15.2% may be used for studying factors that modulate the frequency of VL30 retrotransposition.

To address the potential role of VEGF and/or tinzaparin treatment on VL30 retrotransposition, HC11 cl.17 cells were initially cultured in serum-free medium for 24 h to exclude the presence of serum associated-VEGF. Following serum starvation for 24 h, (a condition used in all subsequent treatments), the retrotransposition frequency of HC11 cl.17 cells was increased from 15.2 to 29% (Fig. 1A). Then, HC11 cl.17 cells were serum-starved in the absence or presence of 3, 6 or 12 ng/ml VEGF for 24 h and their retrotransposition was compared with serum-starved cells without VEGF as the control, VEGF treatment increased retrotransposition in a dose-dependent manner in all cases (Fig. 1A). The level of retrotransposition under serum starvation (29%) increased in the presence of increasing VEGF concentrations (3, 6 and 12 ng/ml) to 30.4, 33.4 and 40.3% corresponding to a net VEGF-dependent induction of 1.4, 4.4 and 11.3%, respectively.

To investigate the action of tinzaparin on VEGF-induced VL30 retrotransposition, serum-starved cells were initially pre-incubated with 2 IU tinzaparin/ml for 4 h, and then treated, in the presence of tinzaparin, with 3, 6 or 12 ng/ml VEGF for 24 h. Compared with those treated with VEGF alone, the retrotransposition frequencies of tinzaparin/VEGF-treated samples were significantly reduced. Specifically, at 3, 6 and 12 ng/ml VEGF, tinzaparin reduced retrotransposition frequency from 30.4 to 28.7%, 33.4 to 24.6%, and 40.3 to 18.3%, respectively, corresponding to a net reduction in retrotransposition frequency of 1.7, 8.8 and 22% (Fig. 1A).

Next, the effects of tinzaparin pre-incubation time on the inhibition of retrotransposition were assessed using treatment times <4 h. HC11 cl.17 cells were pre-incubated with 2 IU tinzaparin/ml for 0.5, 1, 2 or 4 h, and then treated with 12 ng/ml VEGF for 24 h. Using the 39.42% retrotransposition frequency of tinzaparin-untreated cells as the control, tinzaparin reduced VEGF-induced retrotransposition frequency in a time-dependent manner (Fig. 1B). At 0.5, 1, 2 and 4 h, respective retrotransposition frequencies of 36.1, 32.2, 26.3 and 18.7% were recorded, corresponding to a net inhibition of 8.37, 18.27, 33.24 and 52.53%.

In preliminary experiments, a higher concentration of VEGF (i.e. 15 compared with 12 ng/ml) induced an even greater retrotransposition frequency than the 40.3% presented in Fig. 1 (data not shown). It was concluded that: i) 40.3% induced retrotransposition was suitable enough to reliably document the effects of tinzaparin with VEGF treatment; and ii) such a high value (40.3%) was adequate to more accurately assess the effectiveness of tinzaparin, especially at low concentrations (such as the 1 or 2 IU used), and to retain HC11 cell functionality at this level of VEGF-derived retrotransposition, as a higher VEGF concentration may induce many more mutations [since 1% induced retrotransposition equates to an additional ‘retrotransposition burden’ of 1000-fold genome-mutations/cell (33)]. Collectively, these data revealed that tinzaparin was effective in simultaneously reducing serum-starved/VEGF-induced VL30 retrotransposition.

The aforementioned data prompted the investigation of VEGF and/or VEGF/tinzaparin treatment on cell types other than epithelial HC11 mammary progenitors. A previously generated VL30 retrotransposition-positive clone (Pcl.10), derived from mouse NIH3T3 fibroblast cells, was used. Notably, Pcl.10 is a hygromycin B-resistant NIH3T3 fibroblast clone, isolated as HC11 clones, but VL30 retrotransposition events are elicited solely following a stimulus such as the SV40 large T antigen (40), heavy-metal vanadium (VOSO₄) (41) or arsenic (42). Serum starvation for 24 h induced VL30 retrotransposition frequency by 2.4%, compared with non-serum starved Pcl.10 cells. Furthermore, treatment with 12 ng/ml further increased retrotransposition frequency to 7.44% (Fig. S4), corresponding to a net increase of 5.04%. Notably, VEGF concentrations up to 50 ng/ml did not increase retrotransposition frequency further (data not shown). Then, Pcl.10 cells were pre-incubated with 2 IU tinzaparin/ml for 0.5, 1, 2 or 4 h, and then treated with 12 ng/ml VEGF for 24 h. Of note, 7.44% VEGF-induced retrotransposition frequency (used as control), was inhibited in a time-dependent manner to 6.5, 4.75, 3.8 and 2.45%, respectively (Fig. S4). The use of Pcl.10 cells showed that the reduction of VEGF-induced retrotransposition by tinzaparin is not limited to mouse mammary epithelial stem-like HC11 cells.

Serum starvation has been reported to induce the production of reactive oxygen species (ROS) in prostate cancer cell lines (54), and our previous study showed that H₂O₂-derived oxidative stress induces VL30 retrotransposition of NIH3T3 fibroblast Pcl.10 clone cells to a high level (44). Therefore, the effect of oxidative stress on HC11 retrotransposition, and whether it is affected by tinzaparin treatment, was investigated in the present study. VL30 retrotransposition signals activation of a caspase-independent and p53-dependent death pathway (45), while HC11 cells, harboring a mutated p53 gene (55), alleviated cell death allowing the measurement of high retrotransposition frequencies.

In the present study, HC11 cl.17 cells (Fig. S1B) were initially treated with 5–40 pg/ml GO for 24 h, and retrotransposition frequency was measured after a 48 h recovery in normal medium, using non-treated cells as the control. At 10–40 pg/ml GO, the 15.2% nominal retrotransposition frequency of HC11 cl.17 cells was further increased in a concentration-dependent manner, up to ~82.5% at 40 pg/ml (Fig. 2A). Next, considering that 40 pg/ml GO strongly induced VL30 retrotransposition, HC11 cl.17 cells were pre-incubated with 1 or 2 IU/ml tinzaparin for 3 h, then treated with 40 pg/ml GO for 24 h. Pre-incubation with 1 IU- or 2 IU/ml tinzaparin reduced the ~82.5% GO-induced retrotransposition frequency to respective values of 69 and 57%, corresponding to a 13.5 or 25.5% level of inhibition (Fig. 2B). The data indicate that tinzaparin, reducing GO-induced retrotransposition, acts as an anti-oxidant agent.

Our previous study reported that VL30 retrotransposition-positive HC11 cells acquire tumorigenic features such as a high rate of proliferation, induced-EMT phenotype and anchorage-independent mammosphere formation (46). These findings prompted the investigation of tinzaparin on the features of VL30 retrotransposition-positive HC11 cells.

Using both the previously well-characterized HC11 cl.19 for its retrotransposition-induced CSCs and mammosphere formation properties (46), and HC11 cl.15 from the present study, the effect of tinzaparin on HC11 cell proliferation rate was initially examined. Trypsinized cells from these clones were subjected to real-time cell analysis in the absence or presence of 2 IU/ml tinzaparin for up to 80 h, using normal HC11 cells as the control. In the absence of tinzaparin, the continuous proliferation rate of both retrotransposition-positive clones was markedly higher than that of the control cells, but was inhibited to varying degrees in the presence of tinzaparin (Fig. 3A). However, the tinzaparin-induced reduction of proliferation was retained up to 60 h, as previously observed (46), most probably due to the continuous cell proliferation in the absence of culture media replenishment. To quantify the inhibitory effect of tinzaparin on cell proliferation at longer time-points, cells from each clone were cultured in the absence or presence of 2 IU/ml tinzaparin, and counted daily for 5 days. Tinzaparin inhibited the proliferation of both clones in a time-dependent manner, reaching up to ~55 and ~60% for clones 19 and 15, respectively, at day 5 (Fig. 3Ba and b).

VL30 retrotransposition-positive HC11 clones produce CSCs forming anchorage-independent mammospheres (46); therefore their ability to produce cell foci in semi-solid media, and whether this is affected by tinzaparin, was assessed. Trypsinized cells from either normal HC11 or HC11 cl.19 cultures were maintained for three weeks in soft-agar plates containing normal RPMI medium. While normal HC11 cells were not capable of forming foci (Fig. 4, left panel), the respective HC11 cl.19 cells produced large expanding foci (Fig. 4, middle panel). Furthermore, in the presence of 2 IU/ml tinzaparin, the resultant foci were smaller in size (Fig. 4, right panel).

Next, the effect of tinzaparin on mammosphere formation was investigated. HC11 cl.19 cells were seeded into normal culture dishes and their growth was monitored. 10 days after reaching full confluence, multinucleated cells bearing cytoplasmic vacuoles were observed, as well as cell clusters with emerging mammospheres (Fig. 5Aa, arrows). In parallel, examining growth in non-adherent culture dishes for 20 days, the formation of large spheroid cell masses, or mammospheres floating in the culture medium, was apparent (Fig. 5Ab). Next, selected mammospheres transferred into fresh dishes were further cultured in the presence of 2 IU/ml tinzaparin for two or three weeks. After two weeks of tinzaparin administration, substantially reduced mammosphere-size, accompanied by a small number of either disaggregated floating or single mammosphere cells attached to the culture dish, were observed (Fig. 5Ba and b). Notably, after three weeks, the effect of tinzaparin was augmented, as even smaller-sized mammospheres were observed, and the vast majority of single disaggregated mammosphere cells were attached to the dish, bearing a clear mesenchymal phenotype (Fig. 5Bc and d).

Since tinzaparin resulted in mammosphere disaggregation, the potential link between tinzaparin and the modulated retrotransposition of disaggregated cells was investigated. Compared with untreated HC11 cl.19 cells, the retrotrans-position frequency of disaggregated HC11 cl.19 mammosphere-derived mesenchymal-like cells was assessed after treatment with 2 IU tinzaparin for 3 weeks. The retrotransposition frequency was reduced by 49.7% compared with the control, namely from their nominal HC11 cl.19 retrotransposition value of 21.1, to 10.5% (Fig. 5C). Collectively, these data link the inhibitory effect of tinzapapin on VL30 retrotransposion frequency with the inhibition of cellular proliferation, reduction of cell foci size, and the disaggregation of mammospheres of tumorigenic VL30 retrotransposition-positive HC11 cells.

The retrotransposition of autonomous LTR-retrotransposons requires a retrotransposon RNA-intermediate, and the activity of a functionally self-encoded reverse transcriptase (31). To investigate the effect of tinzaparin on the inhibition of VL30 retrotransposition in HC11 cells, semi-quantitative RT-PCR analysis was used to quantity transcription of the retrotransposon RNA-intermediate and/or various enRT gene transcripts.

cDNAs were prepared from normal HC11 cl.19 cells used as control, as well as HC11 cl.19 cells treated with either tinzaparin or VEGF alone, or pre-incubated with tinzaparin and then treated with VEGF for 24 h. Fractionation of PCR products revealed a primary 371 bp band of transcriptionally active VL30s, as well as a smear of enRTs transcripts representing transcriptional active enRT genes, mostly in the range of 1,000-200 bp (Fig. 6A). While 2 IU/ml tinzaparin for 2 h had almost no effect on VL30 RNA expression, the relative expression of enRTs was inhibited ~0.23-fold in comparison to the control (arbitrarily considered as 1-fold). Notably, at 4 h of tinzaparin treatment, the RNA expression of VL30s was inhibited ~1.15-fold, while that of enRTs was further inhibited 0.40-fold. Regarding cell treatment, 12 ng/ml VEGF alone induced RNA expression of VL30s and enRTs of ~1.41- and 1.16-fold. However, 4 h pre-incubation with 2 IU/ml tinzaparin, followed by 12 ng/ml VEGF, inhibited both relative RNA expression of VL30s and enRTs at the level of 0.60- and 0.52-fold, respectively, compared with control levels (Fig. 6A and B). Given that a 4 h tinzaparin pre-incubation simultaneously inhibited the VEGF-induced RNA expression of both VL30s and enRTs, the data indicate that tinzaparin acts rapidly as an anti-retroviral drug inhibiting the necessary VEGF-induced transcripts required for a retrotransposition event.