Targeting antisense mitochondrial ncRNAs inhibits murine melanoma tumor growth and metastasis through reduction in survival and invasion factors

Abstract

We reported that knockdown of the antisense noncoding mitochondrial RNAs (ASncmtRNAs) induces apoptotic death of several human tumor cell lines, but not normal cells, suggesting this approach for selective therapy against different types of cancer. In order to translate these results to a preclinical scenario, we characterized the murine noncoding mitochondrial RNAs (ncmtRNAs) and performed in vivo knockdown in syngeneic murine melanoma models. Mouse ncmtRNAs display structures similar to the human counterparts, including long double-stranded regions arising from the presence of inverted repeats. Knockdown of ASncmtRNAs with specific antisense oligonucleotides (ASO) reduces murine melanoma B16F10 cell proliferation and induces apoptosis in vitro through downregulation of pro-survival and metastasis markers, particularly survivin. For in vivo studies, subcutaneous B16F10 melanoma tumors in C57BL/6 mice were treated systemically with specific and control antisense oligonucleotides (ASO). For metastasis studies, tumors were resected, followed by systemic administration of ASOs and the presence of metastatic nodules in lungs and liver was assessed. Treatment with specific ASO inhibited tumor growth and metastasis after primary tumor resection. In a metastasis-only assay, mice inoculated intravenously with cells and treated with the same ASO displayed reduced number and size of melanoma nodules in the lungs, compared to controls. Our results suggest that ASncmtRNAs could be potent targets for melanoma therapy. To our knowledge, the ASncmtRNAs are the first potential non-nuclear targets for melanoma therapy.

Keywords: antisense therapy; melanoma; metastasis; mitochondria; noncoding RNA.

Figure 1. Deduced structure and expression of the mouse ncmtRNAs

A. Model of the origin of ncmtRNAs. A simplified scheme of mitochondrial transcription shows that both strands of the mtDNA are transcribed bidirectionally. Shown in red is the transcript arising from the light strand (L-strand) promoter (LSP) and in blue, the transcript from the heavy strand (H-strand) promoter HSP2. For simplification, the H-strand transcript arising from HSP1 is not shown. Segments originated from the 16S rRNA gene are then processed to give rise to SncmtRNA and ASncmtRNA. Also shown is the 12S rRNA gene and coding genes for proteins of the NADH dehydrogenase complex (ND1, ND2, ND3, ND4, ND4L, ND5 and ND6), cytochrome oxidase complex (COI, COII and COIII), ATP synthase complex (ATP6 and ATP8) and cytochrome B; the 22 tRNAs are shown in grey. B. Schematic representation of the mSncmtRNA, indicating the size of the loop, the length of the IR and position of oligonucleotides 11S and 11AS used as probes. C. ISH of C2C12 mouse myoblasts and mouse embryonic fibroblasts (MEF) showing a positive signal for the 11AS probe (B), confirming the expression of the mSncmtRNA, and a positive hybridization signal also with the 11S probe (B), corresponding to a putative mASncmtRNA. Bars = 25 μm. D. ISH of sections of normal mouse tissues, testis, intestine (Intes.), skin and seminal vesicles (SV) with probes 11S for mASncmtRNA and 11AS for mSncmtRNA. Hybridization signal was observed with both probes. Bars = 50 μm. E. No ISH signal was observed in normal mouse liver, muscle or brain sections, for probe 11AS (S) or 11S (AS). A parallel section was stained with H & E. Bars = 50 μm. F. Northern blot of total MEF and testis RNA using probes for mSncmtRNA (S) and mASncmtRNAs (mAS1 and mAS2) (see Materials and Methods). EtBr, ethidium bromide-stained gel. G. Schematic representation of mASncmtRNA-1 and −2, indicating the length of the loops and the IRs. ASO-1560S, complementary to both loops, was used for knockdown of these transcripts.

Figure 2. ASK induces death and proliferative index decrease in B16F10 cells

A. ISH of the murine melanoma cell lines B16F10 and B16F0 and renal carcinoma cell line RenCa, using probes 11S and 11AS (Figure 1A). Notice that, in contrast with normal cells, the mASncmtRNAs are downregulated in the three mouse tumor cell lines. Bars = 25 μm. B. B16F10 cells were transfected for 48 h with 200 nM ASO-C, or ASO-1560S, or left untreated (NT). At 48 h post-transfection, ASO-1560S caused massive cell detachment from the substrate, as opposed to controls (NT, ASO-C). STP was included as positive control. Bars = 50 μm. C. Cells were transfected with 200 nM ASO-C or with 100, 150, 200, 250 or 300 nM ASO-1560S, or left untreated for 48 h and cell death was determined by Trypan blue (Tb) exclusion. ASK with 200-300 nM ASO-1560S induces over 60% of cell death, compared to 15-20% in NT and ASO-C controls. D. B16F10 cells transfected with ASO-1560 or ASO-C as in (B) or left untreated for 48 h were collected at different time intervals. ASO-1560S induces a drastic inhibition of cell proliferation compared to NT and ASO-C controls (NT) (*p<0.01). E. RT-PCR of total RNA purified from cells treated as in (B) for 48 h revealed the specific knock-down effect of ASO-1560S on mASncmtRNA-1 (mAS-1) and mASncmtRNA-2 (mAS-2). GAPDH mRNA was used as internal control. M, 100-bp ladder. F, G. Normal mouse cells are not affected by ASK. C2C12 (F) and MEF cells (G) were seeded at 5×10⁴ cells/well in 12-well plates and transfected the next day with ASO-C or ASO-1537S or left untreated (NT). After 48 h, cells were harvested and stained with Tb. A triplicate analysis shows that viability of these two non-tumoral mouse cell types was not affected by ASK. Error bars represent average ± s.d.

Figure 3. Death of B16F10 cells by ASK displays parameters of apoptosis

A. ASK induces dissipation of Δψm. B16F10 cells were transfected with 200 nM ASO-C or ASO-1560S, or left untreated for 24h. A parallel culture was incubated with the uncoupling drug CCCP (see Materials and Methods). Cells were then harvested at 24 h post-transfection, stained with 20 nM TMRM for 15 min and analyzed by flow cytometry. ASK and CCCP induce dissipation of Δψm. B. A triplicate analysis shows that ASO-1560S induces about 55% dissipation of Δψm compared to 95% for CCCP. C. ASK induces release of cytochrome c. B16F10 cells treated as described in (A) or incubated with 5 μm STP as positive control were harvested at 24 h post-transfection. Western blot of the cytosolic fraction (see Materials and Methods) reveals, as compared to controls, a significant release of cytochrome c to the cytosol in cells transfected with ASO-1560S or incubated with STP. D. ASK induces caspase activation. B16F10 cells treated as in (A) for 24h were labeled with FITC-VAD-fmk (see Materials and Methods). Parallel cultures were incubated with 5 μM STP as positive control for apoptosis. Only cells treated with ASO-1560S or STP exhibited green fluorescence due to caspase activation. Bars = 25 μm. E. A triplicate analysis shows that ASK and STP induce over 60% of cells with activated caspases as compared to controls (**p<0.01). Error bars represent average ± s.d. F. ASK induces an increase of sub-G1 fraction. B16F10 cells were treated as in (A), but for 48 h. A parallel culture was incubated with STP as positive control. Cells were then harvested, stained with PI and analyzed by flow cytometry. Only ASO-1560S or STP elicited a significant increase in sub-G1 DNA fraction of 13% and 6%, respectively (*p<0.005). G. ASK induces annexin V-positive cells. B16F10 cells were treated for 24 h as in (A), labeled with annexin V-Alexa Fluor 488 and analyzed by flow cytometry. A triplicate analysis shows that ASO-1560S induced about 45% annexin-positive cells as compared to controls (NT and ASO-C) (*p<0.005).

Figure 4. ASK induces downregulation of survivin

A. B16F10 cells were seeded in 6-well plates at 10⁵ cells/well and transfected the next day with 200 nM ASO-C or ASO-1560S or left untreated (NT). At 24 h, cells were harvested and analyzed by Western blot. Survivin was drastically decreased by ASK, compared to controls (NT and ASO-C). B. A triplicate densitometric analysis of the experiment in (A) indicated that ASK induced around 90% decrease in survivin compared to controls (**p<0.001). C. Quantitative RT-PCR indicated that survivin mRNA is reduced to around 30% by ASK, compared to controls (**p<0.01).

Figure 5. ASK reduces tumorigenic and invasive potential of B16F10 cells

A. ASK inhibits anchorage-independent growth. B16F10 cells were transfected for 48h with 200 nM ASO-C or ASO-1560S or left untreated. After harvesting and counting, 2000 Tb⁻ cells were seeded in soft agar and colony formation was determined after 21 days in culture (see Materials and Methods). Bars = 100 μm. B. A triplicate assay indicated that ASK induced an inhibition of colony formation of over 90% compared to ASO-C (**p<0.005). Error bars represent average ± s.d. C. ASK induces inhibition of stemness. B16F10 cells were transfected as in (A) for 48 h and after harvesting, 4000 Tb⁻ cells were seeded on an agarose-coated plate and cultured for 10-11 days, when spheres >70 μm were scored. Bars = 100 μm. D. A triplicate analysis shows that upon transfection with ASO-1560S, the efficiency of sphere formation was less than 0.03%, compared to 0.35-0.4% in controls (**p<0.005).

Figure 6. ASK reduces invasiveness parameters

A. For matrigel invasion assay, 10⁵ B16F10 cells were transfected with ASO-1560S or ASO-C or left untreated for 48 h. Cells were collected and seeded onto the upper insert. After 24 h, the insert was stained with DAPI and invading cells were scored. B. A triplicate analysis of the experiment in (A) showed that ASK induces a drastic inhibition of B16F10 invasion (**p<0.005). C. B16F10 cells treated as in (A) for 48 h were collected and total lysate was subjected to Western blot with N-cadherin antibody. Representative blot shows a strong decrease in N-cadherin by ASK. D. A triplicate analysis shows a 50-60% downregulation of N-cadherin, compared to controls (**p<0.005). Error bars represent average ± s.d. E. Western blot of β-catenin shows a marked reduction in this protein in 1560S-treated cells. F. A triplicate analysis shows about an 80% reduction of β-catenin, compared to controls (*p<0.002). G. Cells treated as in (A) were collected and total supernatant was subjected to pull-down assay for active Rac-1 and precipitates were analyzed by Western blot. The portion of active Rac-1 was markedly reduced by ASO-1560S. H. A triplicate analysis showed that the proportion of active Rac-1 was reduced to about half of controls, in cells treated with 1560S (**p<0.001).

Figure 7. ASK in vivo inhibits tumor growth and metastasis

A. Scheme of the syngeneic assay. B. Twelve C57BL/6 mice were injected subcutaneously with 100,000 B16F10 cells and at 11-12 days post-cell injection mice exhibited tumors between 700 and 1,000 mm³. The mice were randomized into two groups of six mice with similar tumor volume and, after anesthesia, tumors were surgically removed. Starting day 3 post-surgery, mice received 6 injections of 250 μl of saline containing 100 μg of ASO-C or ASO-1560S, alternating between ip and iv (A). At day 12 post-surgery, mice receiving ASO-C showed tumor relapse and were sacrificed under anesthesia. Tumors, lungs and livers were excised and fixed in formalin. No relapse was observed in the mice receiving ASO-1560S up to day 120 post-surgery, when they were euthanized under anesthesia and lungs and livers were collected and fixed in formalin. The livers C. and lungs D. of mice treated with ASO-C exhibited multiple metastasis nodules of different sizes (red arrows). In contrast, at 120 days post-surgery, livers (C) and lungs (D) were free of metastasis in mice treated with ASO-1560S.

Figure 8. ASK in vivo retards tumor growth through antisense effect

A. ASO-1560S induces tumor growth inhibition. Ten mice were injected subcutaneously with 100,000 B16F10 cells. On day 10 post-cell injection, tumor volumes reached about 100 mm³, mice were randomized into two groups of 6 mice each and were ip injected every other day with 250 μl saline containing 100 μg of ASO-C or ASO-1560S, 6 injections in total. Tumor volumes were monitored with a caliper. The following day after the last injection, tumors were excised and processed to obtain total protein, RNA and tissue sections (see Materials and Methods). B. Representative blots showing that treatment with ASO-1560S inhibits survivin expression. T1-3 tumors from ASO-C-treated mice; T4-6, tumors from ASO-1560S-treated mice. Sur = survivin; β-act = β-actin, used as loading control. C. Densitometry of blots showed that survivin expression in ASO-1560S-treated tumors was about 30% of the control level (*p<0.002). D, E. Levels of mASncmtRNA-1 and −2 were determined by end-point RT-PCR using 18S rRNA as control. The ratio between mASncmtRNAs and 18S rRNA was determined by densitometry. Treatment with ASO-1560S induces a relative inhibition of the levels of mASncmtRNA-1 (**p<0.005) and mASncmtRNA-2 (*p<0.05) compared to ASO-C. Error bars represent average ± s.d. F. Chromogenic TUNEL assay of resected tumors shows augmented apoptosis in tumors from mice treated with ASO-1560S, compared to ASO-C. Bars = 100 μm.

Figure 9. ASK in vivo decreases lung metastasis of B16F10 melanoma induced by tail vein injection of cells

A. Scheme of the assay. Eight C57BL/6 mice were inoculated through the tail vein with 100,000 B16F10 cells. Mice were randomized into two groups of 4 mice each, which received 6 ip injections of 250 μl saline containing 100 μg of ASO-C or ASO-1560S on days 7, 9, 11, 13, 15 and 18 post-cell injection. On day 21 mice were sacrificed and lungs were collected and fixed in formalin. B. Lungs from mice treated with ASO-C exhibited multiple metastasis nodules of different sizes. In contrast, the lungs from ASO-1560S-treated mice treated displayed fewer and smaller nodules. C. Total weight of lungs from ASO-1560S-treated mice was comparable to control mice that had not been inoculated with B16F10 cells (non-inoculated controls; NIC), while the lungs of ASO-C-treated mice displayed an average weight increase of around 2.3 times.