How platelets are produced by megakaryocytes in vivo remains controversial despite more than a century of investigation. Megakaryocytes readily produce proplatelet structures in vitro; however, visualization of platelet release from proplatelets in vivo has remained elusive. We show that within the native prenatal and adult environments, the frequency and rate of proplatelet formation is incompatible with the physiological demands of platelet replacement. We resolve this inconsistency by performing in-depth analysis of plasma membrane budding, a cellular process that has previously been dismissed as a source of platelet production. Our studies demonstrate that membrane budding results in the sustained release of platelets directly into the peripheral circulation during both fetal and adult life without induction of cell death or proplatelet formation. In support of this model, we demonstrate that in mice deficient for NF-E2 (the thrombopoietic master regulator), the absence of membrane budding correlates with failure of in vivo platelet production. Accordingly, we propose that membrane budding, rather than proplatelet formation, supplies the majority of the platelet biomass.

CRISPR technologies have advanced cancer modelling in mice, but CRISPR activation (CRISPRa) methods have not been exploited in this context. We establish a CRISPRa mouse (dCas9a-SAMKI) for inducing gene expression in vivo and in vitro. Using dCas9a-SAMKI primary lymphocytes, we induce B cell restricted genes in T cells and vice versa, demonstrating the power of this system. There are limited models of aggressive double hit lymphoma. Therefore, we transactivate pro-survival BCL-2 in Eµ-MycT/+;dCas9a-SAMKI/+ haematopoietic stem and progenitor cells. Mice transplanted with these cells rapidly develop lymphomas expressing high BCL-2 and MYC. Unlike standard Eµ-Myc lymphomas, BCL-2 expressing lymphomas are highly sensitive to the BCL-2 inhibitor venetoclax. We perform genome-wide activation screens in these lymphoma cells and find a dominant role for the BCL-2 protein A1 in venetoclax resistance. Here we show the potential of our CRISPRa model for mimicking disease and providing insights into resistance mechanisms towards targeted therapies.

The tumour suppressor TP53 (called TRP53 in mice) protects cells from neoplastic transformation by activating diverse cellular processes. While tumour suppression by TRP53-mediated induction of apoptosis, cell cycle arrest, cell senescence and DNA repair has been well characterised, there is so far no example of the loss of a TP53 activated regulator of cellular metabolism promoting tumour development. Using in vivo genome-wide CRISPR knockout screens we identified Nprl3 and Depdc5, encoding components of the GATOR1 complex which inhibits mTORC1, as novel tumour suppressors that inhibit c-MYC-driven lymphomagenesis. In a parallel in vivo CRISPR knockout screen using a focused library of sgRNAs targeting TRP53 binding sites in gene promoters/enhancers, we discovered that loss of a TRP53 binding site in the Nprl3 promoter accelerated c-MYC-driven lymphomagenesis to a similar extent as loss of the Nprl3 gene itself. These findings along with the observations that (i) Nprl3 is upregulated in response to DNA damaging drugs in wild-type (wt) but not TRP53 knockout cells and that (ii) GATOR1 deficient Eμ-Myc lymphomas all retained wt TRP53 function, whereas ~30% of control Eμ-Myc lymphomas had selected for mutations in Trp53, suggested that direct transcriptional inducton of GATOR1 by TRP53 is a critical tumour suppressive mechanism. Additionally, we found that GATOR1 deficient tumours exhibited abnormally increased mTORC1 signalling, which rendered them highly sensitive to the mTORC1 inhibitor rapamycin, both in vitro and in vivo. Collectively, these findings identify the first mechanism by which TRP53 suppresses tumourigenesis by transcriptional activation of a regulator of metabolism and they also reveal a potential bio-marker to predict responses to mTORC1 inhibitors in the clinic.