Sirt1 protects from K-Ras-driven lung carcinogenesis

Abstract

The NAD⁺-dependent deacetylase SIRT1 can be oncogenic or tumor suppressive depending on the tissue. Little is known about the role of SIRT1 in non-small cell lung carcinoma (NSCLC), one of the deadliest cancers, that is frequently associated with mutated K-RAS Therefore, we investigated the effect of SIRT1 on K-RAS-driven lung carcinogenesis. We report that SIRT1 protein levels are downregulated by oncogenic K-RAS in a MEK and PI3K-dependent manner in mouse embryo fibroblasts (MEFs), and in human lung adenocarcinoma cell lines. Furthermore, Sirt1 overexpression in mice delays the appearance of K-Ras^G12V-driven lung adenocarcinomas, reducing the number and size of carcinomas at the time of death and extending survival. Consistently, lower levels of SIRT1 are associated with worse prognosis in human NSCLCs. Mechanistically, analysis of mouse Sirt1-Tg pneumocytes, isolated shortly after K-Ras^G12V activation, reveals that Sirt1 overexpression alters pathways involved in tumor development: proliferation, apoptosis, or extracellular matrix organization. Our work demonstrates a tumor suppressive role of SIRT1 in the development of K-RAS-driven lung adenocarcinomas in mice and humans, suggesting that the SIRT1-K-RAS axis could be a therapeutic target for NSCLCs.

Keywords: K‐RAS; SIRT1; non‐small cell lung carcinoma.

Figure 1. Sirt1 expression in MEFs decreases with time of culture in a K‐Ras‐enhanced manner

1. Western blot against the indicated proteins on samples from WT or Sirt1‐Tg MEFs at 0, 3, and 5 days of culture with 10% FBS. A quantification of the blots (n = 2) is presented in the right panel. Statistical significance of the comparisons between WT and Sirt1‐Tg are represented with *. Statistical significance of the comparisons between the same genotype at day 0 and subsequent days of culture are represented with #.

2. Western blot of the indicated proteins at the indicated times after adeno‐Cre infection.

3. Quantification of the protein levels of Sirt1 after adeno‐Cre infection. Dots represent the average of at least two independent blots. Statistical significance of the comparisons with K‐Ras‐KI; Sirt1‐WT is represented with *. Statistical significance of the comparisons with K‐Ras‐WT; Sirt1‐WT is represented with #.

4. K‐Ras‐KI MEFs infected with the indicated adenoviruses were treated with cycloheximide (CHX) for the indicated times, and the indicated proteins were detected by Western blot. Left panel shows a representative example of n = 5 (GFP) or n = 4 (Cre) different individual clones. Right panel shows the average quantification of Sirt1 protein levels in the four clones.

Data information: Error bars represent the standard error of the mean. Statistical differences were assessed by unpaired two‐tailed Student's t‐test (*P < 0.05; **P < 0.01; ^# P < 0.05; ^## P < 0.01; ^### P < 0.001).Source data are available online for this figure.

Figure EV1. K‐Ras activation in Sirt1‐WT and Sirt1‐Tg MEFs

1. mRNA expression of Sirt1 from WT or Sirt1‐Tg MEFs cultured in 10% FBS for the indicated times. Bars represent the average of at least three independent clones.

2. Sirt1 protein expression in four different clones of immortalized MEFs, after culturing them during the indicated time. Quantification of the band intensity is shown in the right panel.

3. X‐Gal staining of MEFs of the indicated genotypes infected with adeno‐Cre (AdCre) at the indicated times after infection, to detect the LacZ reporter placed in the same polycistronic mRNA with the oncogene K‐Ras^G12V.

4. Growth curve with MEFs of the indicated genotypes after Adeno‐Cre infection.

5. mRNA expression of Sirt1 from three independent replicates of K‐Ras‐KI; Sirt1‐WT or K‐Ras‐KI; Sirt1‐Tg MEFs 4 days after infection with the indicated adenoviruses.

Data information: Error bars indicate standard error of the mean. For comparison between Sirt1‐Tg clones with their corresponding Sirt1‐WT controls, statistical significance was assessed using the unpaired, two‐tailed Student's t‐test. *P < 0.05. For comparison between Sirt1‐Tg clones in different conditions, unpaired two‐tailed Student's t‐test was used. ^# P < 0.05.Source data are available online for this figure.

Figure 2. PI3K and MEK pathways, targets of K‐Ras, participate in the downregulation of Sirt1 expression

1. Left panel: Western blots of the indicated proteins in WT MEFs treated for 16 h with vehicle (DMSO) or the indicated inhibitors. Right panel: quantification of the indicated blots shown in left panel (n = 3).

2. Representative Western blots of the indicated proteins in WT MEFs at day 0 or day 4 of culture, treated with vehicle (DMSO) or the indicated inhibitors for 16 h and with cycloheximide for the indicated times.

3. Quantification of Sirt1 protein levels from n = 3 (DMSO d0, DMSO d4 and MEKi d4) or n = 2 (PI3Ki d4) independent replicates.

4. K‐Ras‐KI MEFs were infected with Adeno‐Cre and treated with MEK inhibitors (Mi) or with PI3K inhibitors (Pi) for 16 h, and the indicated proteins were detected by WB (left panels). Right panel shows a quantification of four independent replicates.

5. Levels of Sirt1 after treatment with the indicated inhibitors for 16 h, relative to the control treatment with DMSO, in the indicated cell lines.

6. Western blot of the indicated proteins from the cell lines where SIRT1 protein levels were shown to respond to MEK and/or PI3K inhibition in (E).

7. Quantification of the three replicates in the Western blots shown in (F).

Data information: Error bars represent the standard error of the mean. Statistical significance was assessed using the unpaired two‐tailed Student's t‐test. ^# P < 0.1; *P < 0.05; **P < 0.01; ***P < 0.001.Source data are available online for this figure.

Figure EV2. Regulation of Sirt1 expression by K‐Ras in MEFs and in human lung tumor cell lines

1. Western blots of the indicated proteins in WT MEFs treated for 16 h with vehicle (DMSO) or the indicated inhibitors.

2. K‐Ras‐KI MEFs were infected with Adeno‐Cre and treated with DMSO (D), MEK inhibitors (Mi) or PI3K inhibitors (Pi) for 16 h, and the indicated proteins were detected by WB.

3. Western bots of the indicated proteins from the indicated cell lines, treated with vehicle (DMSO, D), MEK inhibitor (Mi), or PI3K inhibitor (Pi) for 16 h.

4. Western blots of the indicated proteins from the cell lines where SIRT1 protein levels were shown to respond to MEK and/or PI3K inhibition in Fig 2E.

Source data are available online for this figure.

Figure 3. Increased Sirt1 expression protects from non‐small cell lung carcinoma in mice and humans

1. A

   Kaplan–Meier survival curve of mice of the indicated genotypes injected with 4‐OH tamoxifen at 2‐3 months of age.

2. B

   Kaplan–Meier survival curve of mice from (A) censoring for tumor appearance by μCT scanner.

3. C, D

   Quantification of the number of adenomas (C) and carcinomas (D) per mouse at the time of death by histopathological analysis.

4. E

   Follow‐up of tumor sizes by periodical μCT scanner, starting from the time point when the tumor was first detected.

5. F, G

   (F) Kaplan–Meier curve indicating that lung tumors (including all tumors in the TMA, n = 105) displaying high expression of cytoplasmic SIRT1 (at least 20% of their neoplastic cells positive for SIRT1) show longer progression‐free survival time (***P < 0.0005). (G) Kaplan–Meier curve indicating that lung tumors (including only the adenocarcinomas from the TMA, n = 69) displaying high expression of cytoplasmic SIRT1 show longer progression‐free survival time (*P = 0.029).

Data information: Red lines in panels (C and D) and dots and error bars in panel (E) represent the means and the standard errors of the mean, respectively. Statistical significance was assessed using the log‐rank test (A, B, F, and G) or the unpaired two‐tailed Student's t‐test (C–E). *P < 0.05; ***P < 0.001.

Figure EV3. Sirt1 effects in mouse and human lung carcinoma

1. Quantification of the maximum carcinoma diameter per mouse at the time of death by histopathological analysis. Red lines represent the means and the standard errors of the mean.

2. Kaplan–Meier curve with 105 human lung tumor patients indicating that high expression of cytoplasmic SIRT1 is associated with longer overall survival (**P = 0.003).

Data information: Statistical significance was assessed using the two‐tailed Student's t‐test (A) or the log‐rank test (B). **P < 0.01.

Figure 4. Sirt1 overexpression in mouse pneumocytes modulates the expression of several tumor‐related genes upon K‐Ras‐KI activation

1. A

   Treatment schedule for the isolation of K‐Ras‐KI‐activated pneumocytes. A first phase of 4 weeks of pulse (K‐Ras‐KI activation) by tamoxifen treatment was followed by a chase phase of 2 weeks without tamoxifen for the selection of the remaining K‐Ras‐KI‐activated pneumocytes.

2. B, C

   Representative cytometry heat maps detecting Katushka‐positive cells after only 4 weeks of 4‐OH tamoxifen activation (B), or 4 weeks of 4‐OH tamoxifen pulse + 2 weeks with no tamoxifen treatment chase (C).

3. D, E

   Lists of cancer‐related genes differentially expressed in Sirt1‐Tg pneumocytes compared with Sirt1‐WT pneumocytes in the pulse (D) or in the pulse + chase (E) experiments.

4. F

   Expression in the indicated experiments of the 13 oncogenes (blue dots) and eight tumor suppressors (orange dots) differentially expressed between WT and Sirt1‐Tg pneumocytes and identified by the KEGG platform as belonging to cancer‐related pathways (red bars in Fig EV4D and E). Middle bars indicate the averages, and error bars represent the standard deviations. Statistical significance between oncogenes and tumor suppressors in each experiment was assessed using the unpaired, two‐tailed Student's t‐test. **P < 0.01.

5. G

   Diagram depicting our findings on the functional interaction between oncogenic K‐Ras (K‐Ras^onc) and Sirt1 in lung tumorigenesis.

Figure EV4. Pneumocyte isolation and characterization

1. A

   Strategy for pneumocyte isolation: dissociated lungs were analyzed by cytometry to discard aggregates (first panel); stained with DAPI to exclude dead cells (second panel); and further analyzed by size to discard debris (third panel). Finally, single, alive cells were analyzed for their expression of the lymphocytic marker CD45 or the endothelial marker CD31, and double negatives for CD31/CD45 were considered pneumocytes.

2. B, C

   Representative immunohistochemistry stainings detecting Katushka‐positive cells after only 4 weeks of 4‐OH tamoxifen activation (B) or 4 weeks of 4‐OH tamoxifen pulse + 2 weeks with no tamoxifen treatment chase (C). Size bars represent 200 μm. Arrows indicate Katushka‐positive cells.

3. D, E

   KEGG pathway clustering of differentially expressed genes between the K‐Ras‐KI; Sirt1‐WT and K‐Ras‐KI; Sirt1‐Tg pneumocytes isolated by FACS sorting in the pulse phase (D) or in the pulse + chase phase (E). Red columns represent cancer‐related pathways.