MiR-182-3p targets TRF2 and impairs tumor growth of triple-negative breast cancer

Abstract

The telomeric repeat-binding factor 2 (TRF2) is a telomere-capping protein that plays a key role in the maintenance of telomere structure and function. It is highly expressed in different cancer types, and it contributes to cancer progression. To date, anti-cancer strategies to target TRF2 remain a challenge. Here, we developed a miRNA-based approach to reduce TRF2 expression. By performing a high-throughput luciferase screening of 54 candidate miRNAs, we identified miR-182-3p as a specific and efficient post-transcriptional regulator of TRF2. Ectopic expression of miR-182-3p drastically reduced TRF2 protein levels in a panel of telomerase- or alternative lengthening of telomeres (ALT)-positive cancer cell lines. Moreover, miR-182-3p induced DNA damage at telomeric and pericentromeric sites, eventually leading to strong apoptosis activation. We also observed that treatment with lipid nanoparticles (LNPs) containing miR-182-3p impaired tumor growth in triple-negative breast cancer (TNBC) models, including patient-derived tumor xenografts (PDTXs), without affecting mouse survival or tissue function. Finally, LNPs-miR-182-3p were able to cross the blood-brain barrier and reduce intracranial tumors representing a possible therapeutic option for metastatic brain lesions.

Keywords: TRF2; miR-182-3p; target therapy; telomeres; triple-negative breast cancer.

Figure 1. miR‐182‐3p reduces TRF2 expression in different cancer cell lines

1. A

   Schematic representation of luciferase screening approach. Upper panel shows the four target predictions software used for in silico analysis. Bottom panel indicates the main steps performed in the high‐throughput screening.

2. B

   Upper panel, sequence interaction of miR‐182‐3p with the target site of the wild type 3′UTR of TRF2 in human. Bottom panel, generation of mutant 3′UTR of TRF2 luciferase construct containing the deletion of target site for miR‐182‐3p.

3. C–E

   Luciferase reporter assay in HeLa cells using the synthetic miR‐Control or miR‐182‐3p in combination with the wild type (C) or the mutant 3′UTR of TRF2 construct (D) or the wild type 3′UTR of TRF1 (E).

4. F, G

   Western blotting for TRF2 expression in telomerase‐positive (HeLa, HCT116, MDA‐MB‐231, MDA‐MB‐436) and ALT‐positive (U2‐OS, Saos‐2) cells transiently transfected with miR‐Control or miR‐182‐3p. Upper panel shows the quantification of TRF2 expression. Bottom panel, representative images are shown, actin was used as loading control.

5. H

   U2‐OS cells transiently transfected with the miR‐Control, miR‐182‐3p or miR‐182‐3p inhibitor were assayed by quantitative immunofluorescence for TRF2 3 days post‐transfection. Left panel, representative images. Scale bar: 10 μm. Right panel, quantification of TRF2 fluorescence intensity. a.f.u. arbitrary fluorescence units. N = number of analyzed nuclei. Red bar indicates mean value.

6. I

   U2‐OS cells transfected as described in (H) were assayed by immunofluorescence combined with telomeric FISH. Left panel, representative images of co‐localizations between TRF2 and telomeres (white arrowheads). Scale bar: 10 μm. Right panel, co‐localizations were analyzed using ImageJ software. N = number of analyzed nuclei.

Data information: For (C–G and I), data are shown as mean ± SD. Three independent experiments were performed (n = 3). P values are determined by Student's t‐test; for (H), P values are determined by Mann–Whitney t‐test. Source data are available online for this figure.

Figure EV1. High‐throughput luciferase miRNAs screening identifies miR‐182‐3p as the most efficient miRNA able to target TRF2

1. A

   Results of high‐throughput luciferase screening performed in Hela cells using the wild type 3′UTR‐TRF2 vector in combination with each of the 54 miRNAs selected by in silico analysis. Three days post‐transfection, luciferase ratio (Renilla:Firefly) of each miRNA was calculated, the control miRNA was set “1.” Renilla:Firefly ratios < 1 indicate target specificity of candidate miRNAs for the 3′UTR of TRF2. miRNAs near to the ratio of 0.5 were considered for further analysis. Two biological replicates were performed.

2. B

   HeLa cells transiently transfected with the indicated miRNAs (miR‐Control, miR‐182‐3p, miR‐519e‐5p, miR‐296‐3p) were assayed by western blotting. Upper panel, quantification of TRF2 expression. Bottom panel, representative images of TRF2, TRF1 and RAP1 are shown, actin was used as loading control.

3. C

   Analysis of TRF2 mRNA expression performed by qPCR in four different cancer cell lines (HeLa, MDA‐MB‐231, MDA‐MB‐436, U2‐OS) 3 days post‐transfection with miR‐Control or miR‐182‐3p. The control miRNA was set “1.” Three independent experiments were performed.

4. D, E

   Telomeric ChIP assay in MDA‐MB‐231 (D) and U2‐OS cells (E). Quantification of TRF2 enrichment at telomeric repeats, in the different conditions, is shown in the table under the respective figure. Alu probe and Rabbit IgG were used as negative control for the assay.

Data information: For (A), data are presented as mean values. For (B, C), data are presented as mean values ± SD and Student t‐test was used to calculate statistical significance. Source data are available online for this figure.

Figure EV2. Silencing of TRF2 induces telomeric, pericentromeric and global DNA damage activation

1. A

   MDA‐MB‐231 cells were transiently transfected with the indicated miRNAs or siRNA. The indicated DNA damage markers were assayed by western blotting. Actin was used as loading control.

2. B

   Telomeric DNA FISH performed in MDA‐MB‐231 transiently transfected with the indicated miRNAs. Telomere length was measured by TLF software and indicated as arbitrary fluorescence unit (a.f.u). N = number of analyzed nuclei. Black bar indicates mean value.

3. C

   DNA damage markers were assayed by western blotting in HeLa cells. Actin was used as loading control.

4. D

   Immunofluorescence analysis of γH2AX combined with a telomeric FISH probe (TIFs) was performed in HeLa cells transfected with the indicated miRNAs or siRNAs. Co‐localizations of γH2AX with telomeres are indicated as mean number of TIFs per nucleus.

5. E

   Representative images and enlargements of co‐localizations of experiment described in D.

6. F

   Immunofluorescence analysis of γH2AX combined with a SatIII FISH probe (PIFs) was performed in HeLa cells transfected with the indicated miRNAs or siRNAs. The γH2AX‐positive cells with ≥ 1 PIFs per nucleus were analyzed.

7. G

   Representative images of co‐localizations relative to the experiment described in (F).

8. H, I

   MDA‐MB‐231 and HeLa cells over‐expressing TRF2 or an empty vector (pBabe) were transiently transfected with miR‐Control or miR‐182‐3p. TRF2, pATM and γH2AX expression were assayed by western blotting. Actin was used as loading control.

Data information: For (D) and (F), data are presented as mean values ± SD. Three independent replicates were performed. Scale bar: 10 μm. At least 60 nuclei were analyzed in (D) and (F). A Student t‐test was used to calculate statistical significance. For (B), P values are determined by Mann–Whitney t‐test. All the experiments were performed 3 days post‐transfection with the indicated miRNAs or siRNAs. Source data are available online for this figure.

Figure 2. miR‐182‐3p induces telomeric and pericentromeric DNA damage by TRF2 abrogation

1. Immunofluorescence analysis of γH2AX combined with telomeric FISH (TIFs) was performed in MDA‐MB‐231 cells transfected with the indicated miRNAs or siRNAs. The mean number of TIFs per nucleus was analyzed.

2. Representative images and enlargements of co‐localizations (white arrowheads) relative to the experiment described in (A). Scale bar: 10 μm.

3. Immunofluorescence analysis of γH2AX combined with a SatIII FISH probe (PIFs) was performed in MDA‐MB‐231 cells transfected with the indicated miRNAs or siRNAs. The γH2AX‐positive cells with ≥ 1 PIFs per nucleus were analyzed.

4. Representative images of co‐localizations (white arrowheads) relative to the experiment described in (C). Scale bar: 10 μm.

5. Quantification of TIFs in MDA‐MB‐231 cells over‐expressing TRF2 or an empty vector (pBabe), transfected with indicated miRNAs. The mean number of TIFs per nucleus was quantified.

6. Representative images and enlargements relative to the experiment described in (E). White arrowheads indicate co‐localizations events. Scale bar: 10 μm.

7. Quantification of PIFs in MDA‐MB‐231 cells over‐expressing TRF2 or an empty vector (pBabe), transfected with indicated miRNAs. The γH2AX‐positive cells with ≥ 1 PIFs per nucleus were analyzed.

8. Representative images relative to the experiment described in (G). White arrowheads indicate co‐localizations events. Scale bar: 10 μm.

Data information: For (A, C, E, G) data are shown as mean ± SD. Three independent experiments were performed (n = 3). P values are determined by unpaired two‐tailed t‐test. At least 60 nuclei were analyzed for each experimental condition. All the experiments were performed 3 days post‐transfection with the indicated miRNAs or siRNAs. Source data are available online for this figure.

Figure 3. Ectopic expression of miR‐182‐3p reduces cell growth by apoptosis activation

1. A, B

   MDA‐MB‐436 and MDA‐MB‐231 cells underwent two rounds of transfection with miR‐Control, miR‐182‐3p or miR‐182‐3p inhibitor. Starting from the day of the second transfection, cell confluence was monitored by Incucyte every 24 h up to a maximum of 3 days. The percentage of cell confluence was analyzed.

2. C, D

   Cell number of MDA‐MB‐436 (C) and MDA‐MB‐231 (D) cells and TRF2 expression were analyzed by automatic cell count and by western blotting at the end of the experiment described in (A) and (B). Actin was used as loading control.

3. E

   Two‐dimensional scatter plots of Annexin V analysis performed in MDA‐MB‐436 at the end of the second cycle of transfection with miR‐Control, miR‐182‐3p or miR‐182‐3p inhibitor. Red boxes indicate early and late apoptotic cells.

4. F

   Quantification of Annexin V‐positive cells (%) of experiment described in (E).

5. G

   Two‐dimensional scatter plots of Annexin V analysis performed in MDA‐MB‐231 as described in (E).

6. H

   Quantification of Annexin V‐positive cells (%) of experiment described in (G).

7. I, J

   MDA‐MB‐436 cells over‐expressing TRF2 or an empty vector (pBabe) were transiently transfected with indicated miRNAs and cell count (I) or apoptosis (J) analysis was performed 72 h post‐transfection.

Data information: For (A, B) data are shown as mean ± SEM. For (C, D, F, H, I, J), data are shown as mean ± SD. For (A–D) and (I), three independent experiments were performed (n = 3). P values are determined by unpaired two‐tailed t‐test. For (F), (H) and (J), two different biological replicates were performed. Source data are available online for this figure.

Figure 4. Ectopic expression of miR‐182‐3p reduces cell growth by senescence activation in normal cells

1. A

   Western blotting for TRF2 expression in BJ cells transiently transfected with miR‐Control or miR‐182‐3p. The graph represents the quantification of three independent experiments. Representative images are shown, Actin was used as loading control. Unspecific bands are indicated with (*).

2. B, C

   Mean of γH2AX foci per nucleus was analyzed in BJ cells 72 h post‐transfection with the indicated miRNAs. Representative images of γH2AX foci are shown in (C).

3. D

   Immunofluorescence analysis of γH2AX combined with a telomeric FISH probe (TIFs) was performed in BJ cells 72 h post‐transfection with the indicated miRNAs. Left panel: The mean number of TIFs per nucleus was analyzed. Right panel: Representative images and enlargements of co‐localizations.

4. E

   Cell number of BJ cells was analyzed by automatic cell count at the end of the second round of transfection with miR‐Control or miR‐182‐3p.

5. F

   FACS analysis to evaluate cell cycle progression by Propidium Iodide (PI) staining in BJ cells treated as indicated in (E).

6. G

   β‐Galactosidase assay in BJ cells after two rounds of transfection with mimic miR‐Control or miR‐182‐3p. Left panel: Analysis of β‐galactosidase‐positive cells. Right panel: Representative images.

7. H–J

   IL‐6 (H), CXCL1 (I), IL‐8 (J) factors were analyzed by ELISA to evaluate the senescence‐associated secretory phenotype (SASP) in BJ cells treated as indicated in (G).

Data information: For (A, B, D, E and G–J), a student t‐test was used to calculate statistical significance. Scale bars (10 μm). P values are indicated. Source data are available online for this figure.

Figure EV3. Effects of miR‐182‐3p over‐expression in epithelial breast cancer cells

1. A

   TRF2 and γH2AX expression after two rounds of transfection with the indicated miRNAs, was analyzed by western blotting in MCF10A cells. Actin was used as loading control.

2. B–E

   The mean number of γH2AX foci (B) and TIFs (D) per nucleus were analyzed 72 h post‐transfection with the indicated mimic miRNAs in MCF10A cells. Representative images (C) and (E) are referred to the experiment showed in (B) and (D) respectively.

3. F, G

   Cell confluence (F) of MCF10A was monitored by Incucyte, every 24 h starting from the day of the second transfection, and cell number (G) was counted at the end of experiment (day 4).

4. H–I

   Cell cycle progression analysis by PI staining (H) and cell death analysis by Annexin V assay (I) were performed in MCF10A upon two rounds of transfection with the indicated miRNAs.

5. J

   β‐Galactosidase assay in MCF10A cells after two rounds of transfection with mimic miR‐Control or miR‐182‐3p. Left panel: Analysis of β‐galactosidase‐positive cells. Right panel: Representative images.

Data information: Panels (B, D, F, G, J) data are presented as mean values ± SD. A Student t‐test was used to calculate statistical significance. P values are indicated. Source data are available online for this figure.

Figure 5. LNPs‐miR‐182‐3p treatment inhibits tumor growth in vivo

1. A, B

   MDA‐MB‐231 (A) and MDA‐MB‐436 (B) tumor xenografts were treated with LNPs‐empty, LNPs‐miR‐Control or by LNPs‐miR‐182‐3p when the tumors became palpable. Mice were treated 6 times by intravenous tail vein injections with 20 μg of LNPs‐miR‐Control, LNPs‐miR‐182‐3p or equivalent volume of LNPs‐empty as indicated in the scheduling. The mean of tumor volumes (n = 5 per group) is shown.

2. C, D

   Tumors from mice treated in (A) and (B) were processed to measure miR‐182‐3p expression by TaqMan qPCR.

3. E

   Representative images of IHC analysis of the indicated markers on tumor samples from mice bearing MDA‐MB‐231 human breast cancer xenografts. Scale bar: 50 μm.

4. F

   The histograms show the expression of TRF2, calculated as immunoreactivity score (IRS) by IHC, and the count of positive cells to γH2AX, TUNEL or CD31 staining. The analyses were performed on three mice per group, and the points represent the number of field analyzed for each condition.

5. G, H

   Luminescent MDA‐MB‐436 cells were injected into the brain and monitored by IVIS imaging system. After 1 week from implant, treatment with LNPs‐miR‐Control and LNPs‐miR‐182‐3p was performed as indicated in (A) and (B). Representative images from in vivo (upper panel) or ex‐vivo (bottom panel) brain tumors are shown in (G). Boxplots (H) show the measurement of photons for each brain tumor (n = 5 per group) acquired at the indicated times.

Data information: For (A, B, F), data are shown as mean ± SD. For (C, D, H), the line in the middle of the box plot denotes a median value, the limits of box represent the interquartile range (25^th to 75^th percentiles), while, the whiskers denote the minimum to maximum values. For (A–D) and (H), P values are determined by unpaired two‐tailed t‐test; for (F), P values are determined by Mann–Whitney t‐test. Source data are available online for this figure.

Figure EV4. LNPs‐miR‐182‐3p treatment reaches different organs and reduces TRF2 expression in tumor tissue

1. The organs (brain, liver, kidney) taken from mice, previously engrafted with MDA‐MB‐231 cells and treated with LNPs‐empty, LNPs‐miR‐Control or LNPs‐miR‐182‐3p, were assayed for miR‐182‐3p expression by TaqMan qPCR.

2. Representative images show IHC analysis on tumor samples, from mice bearing MDA‐MB‐436 human breast cancer xenografts, with the indicated markers. Scale bar: 50 μm.

3. The histograms show the expression of TRF2 indicated as immunoreactivity score (IRS) and the percentage of positive cells to γH2AX, TIUNEL or CD31 staining in MDA‐MB‐436 xenografts. Three mice per group were analyzed, the points represent the number of field analyzed for each condition.

Data information: For (A, C), data are presented as mean values ± SD. Statistical significance using unpaired (A) or Mann–Whitney t‐test (C) was calculated. Source data are available online for this figure.

Figure 6. miR‐182‐3p impairs tumor growth by targeting TRF2 in PDTX‐derived tumor cells (PDTCs) and patient‐derived tumor xenografts (PDTXs)

1. A, B

   PDTCs #1 and #2 underwent two rounds of transfection with miR‐Control or miR‐182‐3p. Three days after the second transfection, miR‐182‐3p and TRF2 expression were analyzed by TaqMan qPCR and western blotting, respectively. Actin was used as loading control.

2. C, D

   Left panel, area of each PDTCs was measured by ImageJ. Right panel, representative images are shown. Scale bar: 50 μm. At least 85 3D cells were analyzed for each experimental condition.

3. E

   NSG mice implanted with breast PDTX (#2) were treated with LNPs‐empty, LNPs‐miR‐Control or LNPs‐miR‐182‐3p as indicated in the scheduling. Caliper measurement of tumors was taken at the indicated days. The mean of tumor volumes (n = 5 per group) is shown.

4. F

   miR‐182‐3p expression of tumors from mice treated in (E) was assayed by TaqMan qPCR.

5. G

   Representative images of IHC analysis of the indicated markers from tumors of the experiment showed in (E). Scale bar: 50 μm.

6. H

   The histograms show the expression levels of TRF2 measured as immunoreactivity score (IRS), the percentage of positive cells to γH2AX and TUNEL. The analysis was performed on three mice per group, the points represent the number of field analyzed for each condition.

Data information: For (A–F) and (H), data are shown as mean ± SD. For (A–F), P values are determined by unpaired two‐tailed t‐test; for (H), P values are determined by Mann–Whitney t‐test. For the experiments showed in (A, B) and (C, D) two or three biological replicates were performed, respectively. Source data are available online for this figure.

Figure 7. LNPs‐miR‐182‐3p treatment does not cause toxicity or DNA damage in proliferative organs

1. A

   Representative images of intestine sections from mice previously treated with LNPs‐Empty or LNPs‐miR‐182‐3p. H&E staining (scale bar: 200 μm) and IHC analysis with TRF2 or γH2AX antibodies are shown (scale bar: 50 μm).

2. B, C

   Quantification of TRF2 expression as immunoreactivity score (IRS) (B) and of γH2AX‐positive cells (%) (C) on intestine samples.

3. D

   Representative H&E (scale bar: 200 μm), TRF2 and γH2AX images of skin samples corresponding to LNPs‐Empty or LNPs‐miR‐182‐3p treated animals (scale bar: 50 μm).

4. E, F

   Quantification of TRF2 expression as immunoreactivity score (IRS) (E) and of γH2AX‐positive cells (%) (F) on skin samples.

5. G

   Representative H&E (scale bar: 200 μm), TRF2 and γH2AX images of bone marrow samples corresponding to LNPs‐Empty or LNPs‐miR‐182‐3p treated animals (scale bar: 50 μm).

6. H, I

   Quantification of TRF2 expression as immunoreactivity score (IRS) (H) and of γH2AX‐positive cells (%) (I) on bone marrow samples.

Data information: For (B, C, E, F, H, I), data are shown as mean ± SD. A Mann–Whitney test t‐test was used to calculate statistical significance. Four mice per group were analyzed, the points represent the number of field analyzed for each condition. Source data are available online for this figure.

Figure EV5. LNPs‐miR‐182‐3p treatment does not compromise tissue viability in proliferative and non‐proliferative organs

Proliferative (Intestine, Skin, Bone marrow, Spleen) and non‐proliferative (Brain, Heart, Liver, Kidney, Lung) organs taken from mice, previously treated with LNPs‐empty or LNPs‐miR‐182‐3p, were assayed by hematoxylin and eosin (H&E) staining. Sections from two different mice are shown. Scale bar: 200 μm.