Abstract
Loss-of-function mutations of the cyclic-AMP response element binding protein, binding protein (CREBBP) gene have recently been described at high frequencies across a spectrum of lymphoid malignancies, particularly follicular lymphoma (FL) and diffuse large B-cell lymphoma (DLBCL). The multiple effects of this epigenetic regulator on developmental and homeostatic processes have been extensively studied, however, exactly how CREBBP functions as a tumor suppressor and the reasons for its particular predilection for suppression of lymphoid tumors remains unclear. In addition, for many mature lymphoid malignancies, the existence of cancer stem cells is unproven and their provenance and the initial target cell for transformation a source of ongoing debate. Here we use multiple mouse strains to model loss of Crebbp in different lymphoid compartments to address these questions.
We demonstrate that early loss of Crebbp in hematopoietic stem and progenitor cells (HSPC), through disruption by the Mx1-Cre recombinase, leads to increased development of hematological malignancies, particularly of the B-lymphoid lineage that mimic features of human lymphomas. Theses B-cell malignancies are of long latency and are preceded by significant alterations in the proliferation, self-renewal and differentiation of lymphoid progenitors, allowing hyperproliferative lymphoid progenitors whose differentiation is blocked to accumulate. Using an aberrant surface phenotype that resembles the eventual tumor, we identify a pre-malignant population in the peripheral blood of animals that is often evident many months before any disease characteristics. We demonstrate pre-malignant stem cell characteristics for this population in functional experiments, where it generates high level reconstitution of peripheral blood in transplant recipients, but only gives rise to disease in these animals after a long latency. We also utilize this unique cellular population in longitudinal genome scale analyses (clonality, RNA-Seq, ChIP-Seq and exome sequencing) to document the mechanisms of malignant evolution. Linking the increased rate of mutation we describe to Crebbp loss, we also demonstrate increased DNA damage and an altered DNA-damage response in premalignant lymphoid progenitors.
Importantly, using a Cd19-Cre recombinase that excises only within committed lymphoid cells, we are able to demonstrate that when Crebbp is lost at a later stage of lymphoid development, the marked cellular abnormalities described above are completely lost and the development of tumors is no different from normal (Figure, below). Taken together, these findings define the developmental stage-specific tumor suppressor functions of Crebbp and shed light on the cellular origins and subsequent evolution of lymphoid malignancies. In addition, the altered response to DNA damage that we demonstrate upon loss of Crebbp, allied to the increased exposure to physiological DNA-damage during lymphoid ontogeny offers an explanation for the high incidence of CREBBP mutations in mature lymphoid malignancies.
Left panel, Kaplan Meier graph for Mx1-Cre Crebbp mice with loss of Crebbp in the HSPC compartment demonstrates significantly shorter survival vs WT littermates with intact expression of Crebbp. In contrast, when Crebbp is excised in a later lymphoid compartment through Cd19-Cre mediated recombination, right panel, no difference in survival is noted from WT littermate controls.
Left panel, Kaplan Meier graph for Mx1-Cre Crebbp mice with loss of Crebbp in the HSPC compartment demonstrates significantly shorter survival vs WT littermates with intact expression of Crebbp. In contrast, when Crebbp is excised in a later lymphoid compartment through Cd19-Cre mediated recombination, right panel, no difference in survival is noted from WT littermate controls.
Huntly:Novartis: Speakers Bureau; BMS: Speakers Bureau; Ariad: Speakers Bureau; Pfizer: Speakers Bureau.
