The APL paradigm: discovering, modeling cancer in the mouse, and developing cures

In his early career, Dr. Pandolfi cloned and characterized PML-RARα, the product of the long sought after t(15;17) chromosomal translocation of acute promyelocytic leukemia (APL), and its normal counterpart PML (at that time called myl and subsequently renamed PML). Since then, the Pandolfi lab has deconstructed the genetics and molecular complexity underlying APL, in turn discovering that several molecular subtypes of APL exist, which respond differentially to treatment. The team went on to model these APL subtypes in the mouse, and to optimize APL treatments using faithful APL mouse models. As a result, APL is considered cured and effective combinatorial treatments are available for each APL subtype. We further discovered that the genes of APL (e.g. PML, NPM, PLZF) are frequently implicated in the pathogenesis of human cancer at large, beyond their involvement in APL, by acting as tumor suppressor genes (TSGs). The ability to model cancer in the mouse, to study genes relevant to tumorigenesis both in vitro and in vivo, and to use these model systems to develop novel therapies is the legacy of the APL saga and a distinctive and paradigmatic trait of the research carried out in the lab. Importantly, the lab is still actively working on the TSGs of APL, and the pathogenesis of human leukemia.



Development and Cancer

The Pandolfi lab has a long-standing interest in testing the hypothesis that fundamental developmental processes are deregulated in cancer. The analysis of the genes of APL and their associated proteins proved to be a terrific ground to test this hypothesis, as exemplified by PLZF and BCL6, members of the POZ and Kruppel (POK) family of transcriptional repressors. PLZF is the second most frequent RARα partner in APL associated chromosomal translocations, while BCL6 was identified by virtue of its involvement in chromosomal translocations in Non-Hodgkin’s lymphoma. We have shown that both PLZF and BCL6 play critical roles in development and cancer. Our interest in PLZF and BCL6 has more recently allowed us to identify and characterize POKEMON (for POK, Erythroid, Myeloid ONtogenic factor), also a member of the POK family, as a key player in tumorigenesis and developmental control. Another relevant example is represented by NPM1, also a partner of RARα in APL-associated translocations, which also proved to act as an essential developmental gene. NPM is by now the most frequently mutated gene in acute myeloid leukemia (AML). We have developed several mouse models for Npm1 to dissect its role in development and hemopoiesis. The role of developmental genes and pathways in tumorigenesis represents a major focus of the Pandolfi lab.



Tumor Suppressor genes, signal transduction and protein translation

The elucidation of the molecular function of major tumor suppressor genes (TSGs), such as PML, the APL TSGs, PTEN and p53, and novel ones such as INPP4B has been a long long-standing interest of our lab. This line of investigation is epitomized by our analysis of the PTEN phosphatase: by now the most frequently mutated, deleted, silenced and down-regulated tumor suppressor in the post-p53 era, and a key and rate-limiting enzyme that exerts essential tumor suppressive functions in tumors of various histological nature. PTEN is of great interest to us also because it antagonizes the PI3-kinase pathway, which, in turn, impacts on control of mRNA translation initiation. Regulation of this axis as well as ribosome biogenesis control, as driving forces underlying the neoplastic process represent a main focus of our research. Our analysis of the role of PTEN in tumor suppression led to novel fundamental concepts in tumor biology and genetics and to novel synthetic lethality approaches for cancer treatment (e.g. “pro-senescence” therapy for cancer). More recently, we have focused on the regulatory mechanism underlying PTEN dimerization, plasma membrane recruitment and activation. This work unravels novel therapeutic strategies for the treatment of cancers through PTEN reactivation.



Signaling: Cancer stem cells and metabolism

Most recently, we focused to understanding the contribution of metabolic cues to the maintenance of cancer initiating cells during tumor progression and metastasis. As examples, we have characterized the essential role of PML for the maintenance of the leukemic initiating cell (LIC) pool through the regulation of mTOR, as well as of a novel PML–peroxisome proliferator-activated receptor δ (PPAR-δ)–fatty-acid oxidation (FAO) pathway. We are currently investigating a broader role for PML in the control of lipid metabolism, which is critical for metastasis. Similarly, we have implicated PLZF in the control of mTOR signaling. Studies have also been extended to aberrant metabolism by establishing the pro-oncogenic role of the mutant enzyme isocitrate dehydrogenase-2 (IDH2) and proposing its relevance as a therapeutic target for the treatment of human AML.



Non-coding RNAs and ceRNAs

Non-coding RNAs (ncRNAs) have emerged as a new astonishing component of the mammalian transcriptome. miRNAs, linc-RNAs, pseudogenes and circRNAs vastly outnumber the protein coding mRNA dimension. Additionally, a sizable fraction of non-coding RNAs can also encode for functional polypeptides.  We have also proposed, and are currently testing, a new working hypothesis whereby linear and circular RNAs do cross-talk through competition for shared miRNAs, therefore regulating each other in trans, by acting as “competing endogenous RNAs” (ceRNAs). This in turn attributes a putative and uncharacterized RNA dependent function to protein coding and non-coding linear RNAs, as well as to circRNAs. Importantly, this allows bioinformatic prediction and experimental validation, of which ceRNA might cross-talk with other ceRNAs in a given cell type, on the basis of a new “language” or code, made of shared miRNAs response elements (MREs). We are exploring how these new transformative non-coding RNAs and ceRNAs dimensions are involved in tumorigenesis through in vivo and in vitro analyses and a number of high-throughput screenings.


Ultra-precision medicine and co-clinical approach

Based on the experience of the Pandolfi lab modeling APL and cancer in the mouse in order to develop novel effective treatments, we developed and implemented the “Co-Clinical Trial Project”, a new paradigm for conducting clinical trials concurrently in humans and in mice, This platform utilizes a variety of mouse models, with a focus on genetically engineered mouse models, generated to mimic the spontaneous incidence observed for human cancer. By mirroring human clinical studies in these genetically engineered mouse models (i.e. a Co-Clinical approach) scientists can stratify patients and identify optimal therapies based on both protein coding or non-coding molecular determinants (see below for Non-Coding RNAs). A “Mouse Hospital” concept has been also developed to standardize infrastructures and practices for these co-clinical trials towards the development of “ultra-precision” medicine.

More information regarding the mouse hospital can be found here.