Hepatocellular carcinoma (HCC) is the second leading cause of cancer-related death in the world (1). As a result of advanced disease or poor liver function, few patients with HCC are eligible for surgery: the leading chemotherapeutics agents (sorafenib, doxorubicin, and others) do not offer an increase in the average survival time of 6 months (1–3). Most HCCs develop in the context of liver steatosis, fibrosis and cirrhosis caused by chronic inflammation. These disturbances can be part of a clinical spectrum called non-alcoholic fatty liver disease (NAFLD), or of prolonged alcohol abuse and/or hepatitis viral infection. Regardless of the etiology, disease progression to HCC takes several decades and is thus accrued by aging: in a recent US study enrolling 59,907 the median age at the time of diagnosis of HCC was 62 years (4). Cellular senescence refers to a state of stable cell-cycle arrest combined with the senescent associated secretory phenotype (SASP) (5). Great hopes rely on the so called senolytics, agents that selectively induce removal of senescent cells, and alleviate multiple age-related phenotypes (6). An increased number of senescent cells is associated to age-related tissue degeneration during NAFLD-induced HCC (7, 8). Ogrodnik et al. showed that treatment with a combination of the senolytic drugs dasatinib and quercetin (D+Q) leads to an overall reduction of NAFLD in aged dietary-induced NAFLD mice models (9). Xu et al. recently showed in a ground breaking report that oral administration of D+Q to naturally aged mice alleviated physical dysfunction and increased their survival by 36% (10). D+Q previously showed efficacy also in pre-clinical models of fibrotic pulmonary disease (11). If D+Q have an effect also on the treatment of the more serious HCC condition, at the end of the liver disease inflammatory spectrum, is unknown.

To address this issue, first we tested the separate dose-dependent effects of D or Q on the viability of two well-established HCC cell lines (HepG2, Huh-7) (7, 12–16); we administered D or Q to increased concentrations of D or Q (1 nM−1 mM) to HepG2 and Huh-7 cells and identified 1 μM as the maximal cytostatic dose for either drug, without displaying cytotoxic effects (Figure S1A). Subsequently, we tested the impact of combined D+Q on the cellular senescence induced by the chemotherapeutic agent doxorubicin, according to scheme in Figure 1A. Four experimental groups were considered: (1) control cells; (2) cells treated with doxorubicin (DOX) at 100 nM, a concentration slightly cytotoxic in HCC cells (15, 17, 18). DOX was washed out after 24 h and cells were incubated in normal medium for seven additional days, before analyses; (3) D+Q, cells were treated with 1 μM D + 1 μM Q for 24 h; (4) D+Q, DOX; cells were treated with 100 nM doxorubicin for 24 h. Doxorubicin was washed out after 24 h and cells were incubated in normal medium for six additional days, before treatment with D+Q for additional 24 h (Figure 1A). To assess cellular senescence at the end of the treatments, C12FDG, a fluorogenic substrate for β-galactosidase, was used for detection of senescence associated (SA) β-gal positivity, a bona-fide standard marker of senescent cells. C12FDG retention within the cells was evidenced by the distribution plots obtained from the flow cytometer in the area with stronger green fluorescence intensity (Figure S1B). DOXO induced a 40- and 400-fold increase in SA-β-galactosidase activity in HepG2 and in Huh-7 cells, respectively (Figure 1B). Surprisingly D+Q treatment had no effects in removing β-gal positive cells upon DOX treatment, as well as D+Q alone had no significant effects compared to untreated cells (Figure 1B). Next, we quantified the activation of p16, p21, and γH2A.X, established markers of senescence and of DOX-induced DNA damage, respectively (19, 20), using immunofluorescence. As expected, DOX induced significant increase in the amount of nuclei with more than 5 γH2A.X distinct foci marking DNA lesions, and also in p16 and p21 staining, in Huh-7 cells (Figures 1C,D,S1C) and in HepG2 cells (Figures S2A–C). D+Q had no effect in preventing the activation of these markers of senescence upon DOX treatment (Figures 1C,D, S1C, S2A–C). In parallel we measured the expression levels of key genes involved in cellular senescence (p16, p21) and in SASP (IL-6, IL-8, MMP1, MMP3), using qPCR. DOX increased p21, IL-8, MMP1, MMP3, and TGF-β mRNA levels in both Huh-7 and HepG2 cells (Figures S1D, S2D). Interestingly, IL-6 mRNA was not expressed in HepG2 while it was found increased by DOX treatment in Huh-7 cells (Figure S1D). D+Q was largely ineffective in preventing the activation of senescence/SASP genes in both cell types upon DOX treatment (Figures S1D, S2D). In summary, DOX induced cellular senescence in HCC cells, manifested by the SA-β-gal positivity, flattened nuclear morphology, appearance of senescent markers and SASP gene expression; however D+Q did not exhibit senolytic activity in our in vitro experimental setup involving DOX-induced cellular senescence. To investigate the in vivo antitumor efficacy of doxorubicin, D+Q, or the combined treatment, xenograft studies were performed. Subcutaneous HCC xenografts from Huh-7 cells stably over-expressing a far-red fluorescent protein (eqFP650) were established on the dorsal flank of immunodeficient athymic nu/nu mice, and treated until tumor size in the control/untreated group reached 1,400 mm³ (~23 d post-inoculation). Four experimental groups of balb/c nude mice (n = 11 per group) implanted with Huh-7-eqFP650 were created as it follows: (1) CTL, control mice i.p. injected with vehicle alone (PBS); (2) DOX, mice injected with 4 mg/kg doxorubicin at days 7 and 14 post-implantation; (3) D+Q, mice administered with Dasatinib (D, 5 mg/kg) + Quercetin (Q, 50 mg/kg) by oral gavage, at days 9 and 16 post-implantation; (4) D+Q + DOX, mice injected with 4 mg/kg doxorubicin at days 7 and 14 post-implantation, and simultaneously administered with D+Q by oral gavage, at days 9 and 16 post-implantation (Figure 2A). Tumor volume measurement by caliper and eqFP650 in vivo imaging was performed every 2–3 days until euthanasia. Time-dependent tumor volume growth is illustrated in Figures 2B,C: average tumor volume in mice of group 3 (D+Q) exceeded of 50% the average tumor volume in mice of group 1 (CTL) (p = 0.0252). Treatment of doxorubicin reduced tumor growth of 30% (group 2 vs. group 1, p = 0.0486; Figure 2C). Synergistic treatment of mice with D+Q did not further enhance DOX-induced tumor growth inhibition (Figure 2C). Independent assessment of tumor volume using eqFP650-dependent body fluorescence imaging (Figure S3A) or explanted tumor weight (Figure S3B) at sacrifice largely mirrored the values of tumor growth (Figure 2C). Of note, D+Q treated-tumors showed increased proportion of fibrotic-like tissue upon H/E staining (Figure S3C). Overall these data indicate that D+Q is ineffective in clearing chemotherapy-induced HCC senescent cells as also shown on selected tumors sections stained for the SA-β-gal positivity (Figure 2D). Conversely, D+Q alone displayed pro-tumorigenic effects in vitro and in vivo in the absence of chemotherapy. This calls for concern about the utilization of senolytics and of Q, which is often taken as off-the-counter supplement for its documented anti-oxidant properties, in advanced liver diseases. Moreover, a recent clinical trial on D treatment of patients with advanced liver cancer that cannot be removed by surgery, showed no efficacy and significant side effects¹. It remains to be demonstrated what impact of D+Q may have in dietary/genetic mice models of NAFLD slowly progressing into HCC (21). Whereas, D and/or Q treatment might exert a protective/therapeutic effect against NAFLD at high (9) but not low (22) doses, we demonstrate that these compounds are largely ineffective alone, or in synergy with senescence-inducing chemotherapy, against experimental HCC.

KK, MS, MH, PP, and IP: acquisition, analysis, and interpretation of data. MM and JB: critical revision of the manuscript for important intellectual content. MV conceived and designed the study, obtained funding and drafted the manuscript. All authors critically revised the manuscript and approved its final version.