Fig 2: Stat1- and Cdkn2a-dependent immune control of transplanted RT2-cancers and induction of Ki67-p16Ink4a+ senescent cancer cells.a Individual follow-up of tumour volumes. CD8-depleted mice were subcutaneously (s.c.) engrafted with 1 × 106 RT2-, RT2.Stat1-/--, RT2.Cdkn2a-/-- or RT2.CRISPR-Cdkn2a-cancer cells. Treatment with isotype control mAbs (Ctr) or combined immune checkpoint inhibitors (ICB; anti-PD-L1 and anti-LAG-3) once per week was started when tumours were >3 mm in diameter. Cancer size was measured 2 times per week. Number of mice with regressing tumours and the total number of mice is given in parenthesis; RT2 Ctr N = 9, ICB N = 10, RT2.Stat1-/- Ctr N = 6, ICB N = 8, RT2.Cdkn2a-/- Ctr N = 12, ICB N = 14, RT2.CRISPR-Cdkn2a Ctr N = 5, ICB N = 10. Each cell line was given a different lining. Black lines summarise the results for different treatment groups (as obtained from ANCOVA). p-values examine the question whether the treatment effect was different between two genotypes. Mice were killed either when tumours reached the critical diameter of 15–20 mm or ulcerated, or when mice developed signs of wasting. b Representative triple-staining for the senescence marker p16Ink4a (red) and the proliferation marker Ki67 (blue) and for nuclei (green) of the s.c. tumour of individual mice treated as described in (a). Scale bar 2 µm. c Individual data points showing quantification of p16Ink4a+ (left) or Ki67+ (right) cells. Each data point represents the total of three tumour slides measurements, tumours of three individual mice (described in a) were analysed. In a significance was tested by using unequal variances t-test, p-values examines the treatment effect, comparing the ICB-treated RT2-cancers with each ICB-treated knock-out group.

Fig 3: Stat1-dependent immune control of endogenous RT2-cancers and induction of Ki67-p16Ink4a+ senescent cancer cells.a Survival curves of RT2 or RT2.Stat1-/- mice treated with either isotype control mAbs (Ctr, RT2 N = 9, RT2.Stat1-/- N = 4), with adoptive transfer of T antigen-specific CD4+ TH1 cells (AT, RT2 N = 15, RT2.Stat1-/- N = 9), with immune checkpoint inhibitors (ICB, RT2 N = 12, RT2.Stat1-/- N = 4; anti-PD-L1 and anti-LAG-3) or with ICB/AT (RT2 N = 15, RT2.Stat1-/- N = 10). For each AT experiment we generated a non-cytotoxic Tag-specific TH1 cell line. b Cancer size shown by representative hematoxylin and eosin staining of pancreas of RT2 (upper panel) or RT2.Stat1-/- mice (lower panel) after four weeks of treatment. Mice were treated as described in a. Green lines depict the size of RT2-cancers after treatment. After 4 weeks of treatment ICB/AT-treated mice (N = 6) had a two-fold smaller islet size than Ctr (N = 7). c Time course of blood glucose levels as surrogate marker for the growth of the insulin-producing tumours of RT2 or RT2.Stat1-/- mice (median ± interquartile range) treated as described in (a Ctr, RT2 N = 7, RT2.Stat1-/- N = 4; AT, RT2 N = 10, RT2.Stat1-/- N = 9; ICB, RT2 N = 6, RT2.Stat1-/- N = 4, ICB/AT, RT2 N = 12, RT2.Stat1-/- N = 10). The mice were sacrificed after 4 weeks of treatment for ex vivo analysis. d, e Representative triple-staining for the senescence marker p16Ink4a (red) and the proliferation marker Ki67 (blue) and for nuclei (green) (d), e showing quantification of p16Ink4a+ (left) or Ki67+ (right) cancer cells of individual mice, treated as described in a, each data point represents the total of three tumour slides measurements (Ctr, RT2 N = 9, RT2.Stat1-/- N = 4; AT, RT2 N = 12, RT2.Stat1-/- N = 5; ICB, RT2 N = 9, RT2.Stat1-/- N = 3, ICB/AT, RT2 N = 13, RT2.Stat1-/- N = 7), box plots show the median with 25th and 75th interquartile range (IQR), and whiskers indicate 1.5 × IQR. Significance tested by using Log Rank test (a, left), Fishers exact test (a, right), or two-tailed Mann-Whitney test (c, e). RT2.Stat1-/- mice have been censored after 3.7 weeks of treatment. Scale bars 200 µm (b) or 2 µm (d). Number of mice is given in parenthesis.

Fig 4: IFN-? and p21Cip1-dependent immune control of ?-MYC-induced lymphomas by ICB therapy.a Survival curves of control (Ctr N = 27), or immune checkpoint inhibitor (ICB; anti-CTLA-4 and anti-PD-1 N = 29) treated ?-MYC mice, or of ?-MYC mice treated with ICB and anti-IFN-? (N = 16). b, c Triple-staining for the senescence marker p16Ink4a (red, upper panel) or p21Cip1 (red, lower panel), the proliferation marker Ki67 (blue), and for nuclei (green); scale bar 2 µm; representative pictures from Fig. 8c (b). Box plots with individual data points representing p16Ink4a+ (c, left) or Ki67+ nuclei (c, right) of B cells from Ctr- (N = 7), ICB- (N = 9), ICB and anti-IFN-?-treated ?-MYC mice (N = 7), or ICB-treated ?-MYC.p21Cip1-/- mice (N = 7). Lymph nodes were isolated at similar ages. Each point represents triplicates from one mouse, box plot show the median with 25th and 75th interquartile range (IQR), and whiskers indicate 1.5 × IQR. d ?-MYC mice from the individual treatment groups living at day 200, Ctr N = 13, ICB N = 15, ICB and T-cell-depletion N = 11. e Survival curves of either control (Ctr, N = 14) or ICB-treated ?-MYC.p21Cip1-/- mice (N = 10). Significance tested by using Log Rank test (a, e), two-tailed Mann-Whitney test (c).

Fig 5: Stat1- and Cdkn2a-dependent induction of CIS in RT2-cancer cells, but Stat1- and Cdkn2a-independent induction of apoptosis.a–c Assays were performed with RT2-cancer cells (a, b), RT2.Stat1-/--cancer cells (a), Cdkn2a-deficient RT2-cancer cells (b), or RT2.CRISPR-Ctr- control, or RT2.CRISPR-Cdkn2a-cancer cells (c). For the senescence growth assay, cells were cultured either with medium (Ctr) or with medium containing 100 ng ml-1 IFN-? and 10 ng ml-1 TNF for 96 h, washed and then cultured with medium for another 3–4 days. One representative out of 3 independent experiments was given. SA-ß-gal activity was determined after 96 h of culture with medium (Ctr) or with medium containing 100 ng ml-1 IFN-? and 10 ng ml-1 TNF, data show the mean with SD, RT2 Ctr n = 7, CIS n = 7, RT2.Stat1-/- Ctr n = 6, CIS n = 8 (a), RT2 Ctr n = 6, CIS n = 9, RT2.Cdkn2a-/- Ctr n = 4, CIS n = 5 (b), RT2.CRISPR-Ctr Ctr n = 4, CIS n = 5, RT2.CRISPR-Cdkn2a Ctr n = 4, CIS n = 5 (c). For apoptosis induction, cells were exposed to either medium (Ctr) or etoposide (Eto, 100 µM) or staurosporine (Sta, 0.5 µM) for 24 h and then stained for annexin V. Positive cells were detected by flow cytometry data show the mean with SD (a (including gating strategy), b, c), RT2 n = 4, RT2.Stat1-/- n = 4 (a), RT2 Ctr n = 4, Eto, Stau n = 3, RT2.Cdkn2a-/- n = 3 (b), RT2.CRISPR-Ctr and RT2.CRISPR-Cdkn2a n = 3 (c). Significance tested by using unequal variances t-test.