Therapy-Induced Senescence As an Anti-Cancer and Immune-Stimulatory Strategy

Acute myeloid leukaemia (AML) is the most common type of leukaemia in elderly, for which the current gold standard of treatment is chemotherapy. Recently, it has been observed that AML blasts can activate the senescence program in response to chemotherapy (Therapy Induced Senescence, TIS). Cellular senescence is a stable and terminal state of growth arrest, often caused by nuclear DNA damage, associated with the transcriptional activation of a so-called Senescence Associated Secretory Phenotype (SASP), characterized, among others, by cytokines release, reported to promote immune-surveillance. Here we show that blasts, in response to chemotherapy, accumulate DNA damage and activate the senescence program, that in turn leads to HLA molecules upregulation, making them more prone to be cleared by T-cells.

To evaluate TIS in AML blasts, we started by applying chemotherapy treatment (ARA-C) in six AML cell lines with different p53 status and FAB classification, observing reduction in proliferation rate and activation of DNA damage response pathways in the absence of overt apoptosis. We then quantified Senescence-Associated β-galactosidase (SA-βgal) activity and detected induction of senescence longitudinally with different extent. Accordingly, we observed the same features when applying ARA-C ex-vivo to primary AML samples collected at diagnosis.

To delve deeper into the changes associated with the establishment of senescence in primary blasts, we performed RNA-seq analysis and observed an upregulation of pro-inflammatory genes (including IL1, IL6 and IL8) along with genes involved in immunogenicity. Investigating the biological significance of the transcriptional changes observed, we first reported an increase of HLA molecules on the surface of senescent blasts, as measured by FACS analyses. This observation prompted us to study the interaction between the immune system and senescent blasts, exploiting the Mixed Lymphocyte Reaction (MLR) assay. As expected, we detected a higher T-cell activation of both CD8+ and CD4+ subpopulations, accompanied by an increase in immunological synapses events and in apoptosis induction, when co-culturing chemotherapy treated blasts with T cells.

In order to uncover the molecular mechanisms involved in TIS, we disentangled the role of DNA damage and cell cycle arrest in the phenotype observed comparing AML cell lines treatment with either ARA-C or the cdk4/6 inhibitor, which causes cell cycle arrest without inducing DNA damage. We found that HLA molecules overexpression is linked to the establishment of DNA damage response, however, when comparing acute to chronic ARA-C treatment, we observed that expression levels increased with treatment duration, suggesting that this feature is necessary but not sufficient to increase AML immunogenicity. Next, taking advantage of shRNAs (delivered by lentiviral vectors), we investigated deeper into the role of cell cylcle arrest. By stably knocking down p21, a crucial cell cycle inhibitor, we observed that ARA-C treated blasts had a reduced capacity of activating T-cell. Taken together, these observations point out to a crucial role for senescence in the improved immune-based clearance observed upon ARA-C treatment of blasts. Interestingly, a retrospective analysis showed that a cohort of patients clinically considered "responders" displayed a higher SA-βgal activity, further supporting the idea that senescence establishment in AML may act as a tumour suppressor mechanism.

Overall, our study provides mechanistic insights into the biological and cellular response of AML cells to TIS and presents senescence as a positive mechanism able to promote AML eradication. This opens new lines of research aimed to develop novel therapeutic approaches against AML, exploiting senescence-induced features.