Skip to main content

Role of angiogenesis in leukemia

Final Report Summary - LEUKEMIA AND VEGF (Role of angiogenesis in leukemia)

Starting from the observation that high concentrations of the pro-angiogenic factor VEGF are commonly found in the bone marrow of leukemia patients, we hypothesised that overexpression or activation of the transcription factor HIF-1alpha may drive high VEGF expression and bone marrow neo-angiogenesis in leukemia. Hypoxia-inducible transcription factors (HIFs) are the main regulators of cellular and systemic adaptation to hypoxia, and are often upregulated in solid tumours because of intra-tumoural hypoxia and activation of specific oncogenic pathways. In solid tumours, HIF activation triggers a complex series of adaptive responses that range from the induction of anaerobic metabolism, cell migration and neo-angiogenesis, to promoting maintenance of cancer stem cells. In leukemia and lymphoma, the role of HIF factors is less studied: only recently few studies have shown that HIF-1aplha is upregulated in certain leukemic diseases (acute and chronic lymphocytic leukemia), where it may regulate neo-angiogenic processes, or that it is overexpressed in specific cell populations, such as leukemia stem cells (in lymphoma and some acute myeloid leukemias), where it regulates stem cell maintenance.

With our project we had aimed at the following:

- elucidating the cause of VEGF expression in acute myeloid leukemia and chronic myeloid leukemia;
- studying the effect of HIF-1alpha and VEGF overexpression on normal hematopoiesis;
- studying the effect of HIF-1alpha and VEGF overexpression on leukemia onset and progression.

For aim 1, we hypothesised that high VEGF expression in myeloid leukemia is caused by the accumulation or activation of the transcription factor HIF-1alpha and we proposed to assess whether increased accumulation or activation of HIF-1alpha is evident in in human leukemia clinical samples as well as in mice affected by myeloid leukemia.

For aim 2, we hypothesised that forced expression of a stable form of HIF-1alpha would increase the number of hematopoietic stem progenitor cells in the bone marrow, cause an increase in bone marrow microvessel density and perhaps induce leukemia.

For aim 3, we had proposed to provide genetic proof that expression of HIF-1alpha by leukemic blasts increases bone marrow angiogenesis and cooperates with known oncogenes in leukemogenesis by overexpressing or silencing HIF-1alpha in mouse models of myeloid leukemia.

Upon evaluation of HIF-1alpha expression and activation on a panel of myeloid leukemia cell lines, in these four years we have worked predominantly on acute myeloid leukemia, with a focus on acute promyelocytic leukemia (APL), because activation of HIF-1alpha was found particularly elevated in cell lines representative of this disease.

A number of attempts aimed at evaluating the expression of HIF-1alpha in leukemia patients in collaboration with the pathology and leukemia units of San Raffaele Hospital failed because of technical problems with bone marrow samples. Therefore we have focused on studying cell lines and animal models, where we reached a series of important conclusions that we have later confirmed through in silico analysis of published microarray data from leukemia patients.

At the molecular levels, we found that the oncogenic fusion protein of APL, PML-RARalpha, acts as a HIF-transcriptional co-activator through physical and functional interaction with HIF-1alpha, thus leading to the upregulation of HIF-dependent gene. Significantly, in silico analysis of publicly available gene expression data from APL patients revealed that HIF-dependent signatures are over-represented in APL promyelocytes as compared to normal promyelocytes. Therefore, although a direct measurement of HIF-1alpha levels in APL patients was not possible, these data confirmed that activation of HIF-dependent pathways is relevant to the pathophysiology of APL. Moreover, as molecular data obtained in our laboratory indicate that in leukemia HIF-1alpha levels per se may not be predictive of its activation, in the future similar in silico gene expression profiling will be applied to other myeloid leukemias.

Building up on these initial data, a series of experiments were performed both in vitro and in vivo, in cell lines and in different mouse models, to address the physiological significance of the functional interaction of PML-RARalpha with HIF-1alpha. Because HIF-1alpha was found hyperactivated rather than simply overexpressed in APL cells, we studied the role of HIF-1alpha in APL through silencing strategies or by inhibition with pharmacological agents, rather than by forcing its expression. Firstly, by analysing the effect of specific silencing or pharmacological inhibition of HIF-1alpha in vitro, in different APL cell lines, and in vivo in xenograft experiments, we found that PML-RARalpha exploits the transcriptional repertoire of HIF-1alpha to foster leukemia progression at multiple levels: by inducing spontaneous and chemokine-directed cell migration, by up-regulating VEGF and promoting neo-angiogenesis in leukemia-colonised bone marrow, and by regulating maintenance of colony-forming leukemic cells. To substantiate these data in a more physiological setting, we created a mouse model of APL by transducing bone marrow mononucleated cells with PML-RARalpha followed by transplantation into syngeneic mice. With this protocol, a leukemia resembling human APL develops four to six months after transplantation, and leukemic cells propagate the disease by transplantation into syngeneic mice, which maintain features of human APL both as disease composition and progression, as well as in their sensitivity to APL-specific pharmaceutical compounds (e.g. retinoic acid). With this model we further demonstrated that oligonucleotide strategies inhibiting HIF-1alpha slowed leukemia progression by inhibiting cell migration, neo-angiogenesis and colony-formation. Also, we performed a series of pre-clinical trials with clinically relevant pharmacological agents with HIF-inhibitory activity that are currently used in clinical trials for solid tumours. These latest studies allowed us to demonstrate that HIF inhibitors blunt APL leukemia progression when used as single agents, but also specifically and exquisitely synergise with retinoic acid treatment towards eradicating leukemia stem cells and preventing leukemia relapse.

Because of our use of therapeutically relevant compounds currently being tested in clinical trials, our studies will be of great socio-economic interest given their immediate applicability for the treatment of leukemia patients. Moreover, these studies lay the basis for addressing the role and the effect of inhibiting of HIF factors also in other hematological malignancies.