1st Annual Workshop on Systems Biology: Tumor Dormancy


Keynote Speaker Abstracts


Monday, July 25, 2011 Tuesday, July 26, 2011
Wednesday, July 27, 2011
 
 

Julio A. Aguirre-Ghiso
Departments of Medicine and Department of Otolaryngology, Mount Sinai School of Medicine, New York, New York, U.S.A.
 
The target organ microenvironment and stress signaling as determinants of disseminated tumor cell dormancy.
 
The mechanisms driving dormancy of disseminated tumor cells (DTCs) remain largely unknown. Here we explored whether the target organ microenvironment can control the timing of DTC dormancy. HNSCC HEp3-GFP cells spontaneously disseminate to lungs and bone marrow (BM). A dormancy period of ~2 weeks precedes aggressive expansion of lung DTCs (Lu-HEp3), which always (100%) proliferated in culture and displayed an [ERK/p38]high ratio. In contrast, bone marrow DTCs (BM-HEp3) in vivo persisted without developing overt metastasis and only 25% of BM-HEp3 DTCs proliferated when place in culture. Here we reveal that BM-derived DTCs displayed an [ERK/p38] low ratio, induction of the key dormancy transcription factors (TFs) BHLHB3, p53 and NR2F1 and autophagy genes ATG6, ATG7 and ATG8. Upon re-injection in vivo Lu-HEp3 cells were always tumorigenic, while BM-HEp3 cells underwent a dormancy phase and those that eventually grew displayed an [ERK/p38]high ratio and silenced the dormancy TFs. Importantly, systemic inhibition of p38a/b with SB203580 dramatically accelerated lung metastasis and now DTC, micro- and macro-metastasis were detectable in sites where they are usually a rarity such as liver, spleen and spinal fluid. We further identified TGF-b2 as a potential mediator of the reprogramming that leads to a [ERK/p38]low ratio and dormancy gene induction. We also show that conditioned media from bone marrow but not lung whole tissue primary cultures can induce a [ERK/p38]low ratio. Suggesting that a microenvironment specific signal might dictate DTC dormancy vs. proliferation in different target organs. In an attempt to explore the relevance of our findings in human disease we applied the dormancy gene signature identified downstream of a [ERK/p38]low ratio to both breast cancer cell line expression data as well as four published clinical studies of primary breast cancers. We find that estrogen receptor (ER) positive breast cell lines and primary tumors have significantly higher dormancy signature scores (P<0.0000001) than ER- cell lines and tumors. In addition, a stratified analysis combining all ER+ tumors in four studies indicated 2.3 times higher hazard of recurrence among patients whose tumors had low dormancy scores compared to those whose tumors had high dormancy scores (p < 0.000001). The difference was shown in all four individual studies. This analysis suggests that disseminated ER positive tumor cells carrying a dormancy signature are more likely to undergo a prolonged dormancy phase before resuming metastatic growth. We propose that stress-induced quiescence and autophagy contribute to DTC dormancy and that the BM is a dormancy-instructive microenvironment. We further propose that gene signatures identified in this dormancy model might provide information on the progression of breast cancer patients, specifically the ability of the residual disease to remain for long periods asymptomatic and presumably dormant.
 

[top]
 

Nava Almog
Center of Cancer Systems Biology, Steward St. Elizabeth's Medical Center, Tufts University School of Medicine, Boston, Massachusetts, U.S.A.
 
Genes and regulatory pathways involved in persistence of dormant micro-tumors.
 
Tumor dormancy is a stage in tumor progression in which tumors are kept occult and asymptomatic for a prolonged period of time. It is present as one of the earliest stages in tumor development, as micro-metastasis in distant organs, and as minimal residual disease left after surgical removal or treatment of primary tumors and, therefore, is a more common feature of cancer than appreciated. Dormant tumors are usually at a size of a few millimeters in diameter and can switch to become fast-growing, clinically apparent and potentially lethal. Understanding the underlying mechanisms of tumor dormancy, therefore, could have significant implications in the prevention and treatment of cancer.
 
We have previously established in vivo models of human breast cancer, glioblastoma, osteosarcoma, and liposarcoma dormancy in immunocompromised (SCID) mice and showed that in these models, tumor dormancy is associated with impaired angiogenic potential. For each tumor type (glioblastoma, osteosarcoma, breast carcinoma and liposarcoma) we currently have a pair of 2 clones: One that generates dormant tumors and one that generates fast growing tumors. Using these models, we have shown that non-angiogenic, dormant microscopic tumors reside in mice for a long period (>90 days) until they switch to become fast-growing angiogenic tumors. We used genome-wide expression profiling assays to determine the consensus signature of human tumor dormancy. We identified genes that are differentially expressed between dormant and fast growing tumors, regardless of tumor type and characterized common tumor dormancy associated genes.
 
It is becoming well accepted that microRNAs play critical roles in cancer. MicroRNAs repress expression of target genes and can act as either oncogenes or tumor suppressors in tumor development. It is estimated that one microRNA could regulate gene expression of multiple target genes and therefore act as 'master regulator' of gene expression. We have recently compared microRNA expression profiles in our in-vivo models of human tumor dormancy and identified a consensus microRNA signature of dormant tumors. We show that dormant tumors undergo a stable microRNA switch during their transition to the fast-growing angiogenic phenotype and that loss of DmiR expression was the prevailing mechanism that governed the switch of dormant tumors to fast-growth. Identifying the molecular regulators of tumor dormancy may provide molecular instructions to address the unmet medical need for blocking tumor progression in an early stage and developing novel early tumor dormancy biomarkers.  

[top]
 

Heiko Enderling
Center of Cancer Systems Biology, Steward St. Elizabeth's Medical Center, Tufts University School of Medicine, Boston, Massachusetts, U.S.A.
 
Cancer stem cells and tumor dormancy.
 
Cancer stem cells possess the capacity for self-renewal and are responsible for driving tumor resistance, aggressiveness as well as the frequent recurrences seen after apparently eradicative tumor treatment. The recent realization that these cells may also comprise but a minority of a cancer cell population raises the prospect that the non-stem compartment contributes negligibly to tumor progression. However, the frequent clinical observation of locally recurring cancers displaying more aggressive growth dynamics than the original primary tumor, coupled with the widespread observation of indolent tumors among asymptomatic adults, together suggest that the non-stem cells targeted by treatment may compete with a resistant, aggressive subpopulation of cancer stem cells to prevent their enrichment beyond a small fraction. We develop an agent-based model of cancer stem cell and non-stem cancer cell interactions, and use this model to simulate early tumor growth dynamics. We discuss that an environmentally independent dormant state is an inevitable early tumor progression bottleneck for a large range of biologically realistic cell kinetic parameters. When intrinsic cell kinetics combine in unexpected manner, escape to tumor progression occurs as morphologically distinct self-metastatic expansion of multiple self-limited tumor clones.
 

[top]
 

Dean Felsher
Department of Medicine, Stanford University School of Medicine, Stanford, California, U.S.A.
 
Modeling tumor dormancy: Possible role of stem cell renewal programs.
 
The targeted inactivation of oncogenes can elicit sustained tumor regression, associated with the phenomenon of oncogene addiction. We have utilized conditional transgenic mouse models of MYC-induced hematopoietic malignancies, hepatocellular carcinoma and osteogenic sarcoma to define the mechanisms of oncogene addiction. Through these conditional transgenic mouse models, we have gleaned some insight into the mechanisms of oncogene addiction illustrating that upon oncogene inactivation tumors undergo proliferative arrest, apoptosis, differentiation and/or senescence. The specific consequences of oncogene addiction depend upon both cellular and genetic context and both tumor cell intrinsic and host-dependent mechanisms appear to be critical. Our results illustrate that oncogene addiction involves complex interactions between the host and tumor microenvironment may play a critical role in the mechanism by which oncogene inactivation elicits tumor regression. One apparently convergent theme is that oncogene inactivation is often accompanied by the uncovering of stem cell like properties of tumor cells, including the differentiation into sometimes multiple cellular lineages and the permanent loss of self-renewal programs through cellular senescence. We have used a combination of quantitative imaging and in situ measurements of signaling, apoptosis and cell cycle markers to model oncogene addiction. Our results suggest that we can predict oncogene addiction and this is defined by whether a tumor undergoes proliferative arrest and apoptosis, the conversion into a quiescent dormant state or the terminal differentiation into senescent cells.
 

[top]
 

Philip Hahnfeldt
Center of Cancer Systems Biology, Steward St. Elizabeth's Medical Center, Tufts University School of Medicine, Boston, Massachusetts, U.S.A.
 
A host support niche as a control point for tumor dormancy -- Implications for tumor development and beyond.
 
One common example of a "dormant" tumor is a pre-vascular nodule, with size defined by the oxygen diffusion distance, which has not developed the capacity to induce neovascularization to enable continued growth. While the concept of balanced pro- and antiangiogenic influences has been used to describe this state, all we actually know in this case is that tumor-derived induction of host support is insufficient to expand that support to levels that overcome the local diffusion limitation. A less recognized, but potentially even more interesting, state of dormancy unfolds when a tumor that has already traversed this initial impediment to growth re-establishes an equilibrium condition where stimulated host vascular support is just adequate to sustain the current tumor size. Tumors in this state can be of any size, and can present as dormant non-progressing tumor lesions, e.g. of breast, prostate, thyroid etc. The existence of "post-vascular" dormant tumors illustrates an essential concept in tumor progression growth dynamics – the potential to regulate tumor size through control of the tumor support niche itself. The consequence of such cross-compartment, tumor-niche control is that tumors may naturally grow as if anticipating a size limit defined by a theoretical point of dormancy. We extend this reasoning to show that tumors may share with normal tissues and organs a cross-compartment size control mechanism, except that tumors have typically elevated the control point to sizes that are inconsistent with host viability. In this respect, we suggest post-vascular tumor dormancy is more a realization of a governing principle in mass control than an anomalous feature of tumor growth. Evidence from tumor xenograft data in mice subject to antiangiogenic treatment is reviewed, and compared to organ control data and data from breast cancer patients, suggesting a common reciprocal growth control program may be operative throughout. Additionally, the cancer findings point to a robust tumor control dynamic that not only overshadows tumor type and inter-patient diversity, but also appears to recapitulate early mammalian and avian development.
 

[top]
 

Stefano Indraccolo
Instituto Oncologico Veneto-IRCCS, Padova, Italy
 
Insights into the regulation of tumor dormancy by angiogenesis in experimental tumors.
 
While it is established that an angiogenic switch marks escape from tumor dormancy in xenograft models (Indraccolo S. et al. PNAS 2006), the molecular pathways involved in the control of tumor cell proliferation or survival by angiogenesis remain substantially uncharted. In the first part of my talk, I'll present data supporting the hypothesis that signals stemming from angiogenic endothelial cells regulate the behaviour of dormant cancer cells. Recently, we demonstrated that the Notch ligand Dll4, induced by angiogenic factors in endothelial cells, triggers Notch3 activation in neighbouring tumor cells and promotes a tumorigenic phenotype (Indraccolo S. et al. Cancer Res 2009). We further dissected molecular events downstream of Notch in T acute lymphoblastic leukemia (T-ALL) xenografts and observed that MKP-1 levels – a broadly expressed phosphatase – are controlled by Notch3 by regulation of protein ubiquitination and stability but not MKP-1 gene expression (Masiero M. et al. Leukemia 2011). Notch3 and MKP-1 levels are consistently up-regulated in aggressive compared to dormant tumors, and this is accompanied by opposite variations in the levels of active p38, a canonical MKP-1 target. A good correlation between Notch3 and MKP-1 levels was observed in T-ALL primary samples and in a panel of T-ALL cell lines. Silencing Notch3 by RNA interference or g-secretase treatment, or stimulation of Notch3 by the Dll4 ligand substantially reduced MKP-1 levels in T-ALL cells in vitro. Concordant results have been obtained by neutralizing Dll4 both in a co-colture system in vitro and in vivo. Attenuation of MKP-1 levels by shRNA did not affect proliferation, whereas it significantly increased T-ALL cell death following drug treatment or serum starvation. Importantly, engraftment of MKP-1 deficient T-ALL cells in immunodeficient mice was markedly impaired compared to controls. Altogether, these results elucidate a novel angiogenesis-driven mechanism involving the Notch and MAPK pathways that controls survival of T-ALL cells and tumor dormancy.
 
In the second part of my talk, I will discuss the possibility of achieving tumor dormancy by counteracting angiogenesis in established tumors. With few exceptions (such as endostatin), currently available anti-angiogenic drugs – mainly targeting the VEGF pathway – delay tumor growth but they do not cause tumor dormancy and indeed rapid tumor regrowth is observed in the off-therapy period. In experimental tumors, anti-VEGF drugs typically prune the newly formed vasculature, thus reducing microvessel density, blood flow and perfusion and eventually increasing the level of intratumoral hypoxia. Although in patients mechanisms might be more complex, considering that VEGF neutralization has therapeutic efficacy mainly when combined with conventional chemotherapy, it is currently held that following an initial time window of "vascular normalization", regression of tumor vasculature occurs leading to increased hypoxia. Less well understood is whether hypoxia eventually leads to tumor starvation. On one hand, hypoxia can exert both anti-proliferative effects or induce cancer cell death depending on specific genetic features of tumor cells. On the other hand, hypoxia can also favor invasion and metastasis, as an evasive mechanism to anti-angiogenic therapy. Although it is usually granted that anti-angiogenic drugs cut oxygen supply into tumors, surprisingly less is known about other metabolic perturbations induced by VEGF blockade in the tumor microenvironment. I will present data showing that anti-VEGF therapy causes a dramatic depletion of glucose and an exhaustion of ATP levels in tumors, although glucose uptake is maintained. Additionally, we found that the central metabolic protein kinase AMPK – a cellular sensor of ATP levels that supports cell viability in response to energy stress – is activated by anti-VEGF therapy in experimental tumors and it has a key role in attenuating cell proliferation and induction of sustained tumor regression (Nardo G. et al. Cancer Res, in press). Taken together, our findings reveal functional links between the Warburg effect and the AMPK pathway with induction of tumor dormancy following VEGF neutralization in tumor xenograft models.
 

[top]
 

Bruno Quesnel
INSERM, Institut pour la Recherche sur le Cancer de Lille, (IRCL) Lille, France
 
Tumor dormancy: Long-term survival in a hostile environment.
 
The role of the immune system in tumor dormancy is now well established, as prior immunization with tumor cells has been shown to induce tumor dormancy in immuno-competent hosts in several experimental models. Equilibrium between the immune response and tumor cells leads to long-term tumor dormancy. This equilibrium is also observed early in tumor development, and adaptive immunity may help contain tumor outgrowth. Immune responses may target tumor cells themselves or the tumor microenvironment through direct cell killing by cytotoxic T cells and NK cells. In addition, recent data show that CD4 T cells may also control dormant tumor cells by limiting angiogenesis or inducing cellular senescence through mechanisms that remain unclear. After variable periods of time, however, tumor dormancy ends and the disease progresses. As the immune response remains active, the tumor cells presumably escape dormancy by becoming resistant. Due to the extreme difficulty of isolating dormant tumor cells from patients, the mechanisms underlying this resistance are poorly understood; however, experimental models have shown that dormant tumor cells may over-express B7-H1 and B7.1, thereby inhibiting CTL-mediated lysis. TLR ligands and IFN-gamma induce B7-H1 expression, suggesting that chronic infections and inflammatory states may help dormant tumor cells to escape T cell responses. Dormant tumor cells may also resist apoptosis by deregulation of survival pathways such as JAK/STAT and AKT/mTOR or by paracrine production of cytokines. Importantly, these mechanisms of immunoevasion may also lead to cross-resistance to various anti-cancer agents, suggesting that tumor dormancy itself may limit treatment efficacy. The presence of immuno-escape mechanisms in tumor cells from relapsing patients also suggests that the immune equilibrium that maintained dormancy had broken down in these patients. Identification of the mechanisms underlying the loss of immune equilibrium would offer new insight and potentially suggest new therapies that could restore a favorable immune balance and thus mediate the clearance of residual disease. This effect could be achieved through the use of targeted drugs, such as kinase inhibitors that specifically inhibit the expression of immunoescape molecules at the cancer cell surface, that are delivered continuously during the period of complete remission.
 

[top]
 

Tobias Schatton
Brigham and Women's Hospital, Children's Hospital Boston, Harvard Medical School, Boston, Massachusetts, U.S.A.
 
Tumor dormancy: How cancer stem cells outsmart the immune system.
 
Increasing evidence suggests that tumor dormancy represents an important mechanism underlying the observed failure of current therapeutic modalities to fully eradicate cancers. In addition to its more established role in maintaining minimal residual disease after treatment, dormancy might also critically contribute to early stages of tumor initiation and the formation of micrometastatic foci. There are striking parallels between the concept of tumor dormancy and the cancer stem cell (CSC) theory of tumor development. For instance, the CSC hypothesis similarly predicts that a subset of self-renewing cancer cells – that is CSCs – is responsible for tumor initiation, bears the preferential ability to survive tumor therapy, and persists long-term to ultimately cause delayed cancer recurrence and metastatic progression. Additionally, many of the biological mechanisms involved in controlling the tumor dormant state can also govern CSC behavior, including cell cycle modifications, alteration of angiogenic processes, and modulation of antitumor immune responses. In fact, quiescence and immune escape are emerging hallmark features of at least some CSCs, indicating significant overlap between dormant cancer populations and CSCs. Here, we will crucially dissect the potential role of CSCs in the induction and maintenance of tumor dormancy, with a focus on CSC interactions with the immune system. We will elucidate how recently uncovered CSC immunological functions might allow for tumor immune evasion and for regulation of cancer expansion, relapse, and progression. Accordingly, we believe that defining CSC immunobiological pathways could help critically advance our understanding of the cellular events mediating tumor dormancy. Ultimately, such knowledge might be useful in generating novel therapeutics that efficiently target both metabolically active and dormant CSCs to minimize the risk of tumor recurrence.
 

[top]
 

Kathleen Wilkie
Center of Cancer Systems Biology, Steward St. Elizabeth's Medical Center, Tufts University School of Medicine, Boston, Massachusetts, U.S.A.
 
Mathematical models of tumor-immune interactions.
 
The physical presence and activities of cancer cells can elicit an immune response in the host. Although the majority of models presume the immune response acts to suppress tumor growth, it is becoming clear that the immune response can be both stimulatory and inhibitory. The interplay between these competing influences has complex implications for tumor development and cancer dormancy. To explore these, we have developed a two-compartment model consisting of a population of cancer cells and a population of immune cells. Cancer cells, innate immune cells (such as platelets, dendritic cells, macrophages, and natural killer cells) and adaptive immune cells (such as T and B lymphocytes) communicate with each other through cytokine and chemokine production which controls and shapes tumor growth. The cumulative result of the interactions of these diverse cells determines whether a net tumor-promoting or tumor-inhibiting effect occurs, and it is this holistic response that we attempt to capture in our model. Reviewed here are some of the commonly used mathematical models of tumor-immune interactions. As will be seen, these models tend to focus on single immune cell types and their specific function in cancer cell lysis. Our model, on the other hand, combines the effects of all immune cell types, general principles of self-limited logistic growth, and the physical process of inflammation into one quantitative model setting. Thus, it is well-positioned to predict immunomodulation of tumor growth, and to assist in the design of novel treatment approaches that exploit immune response to improve tumor suppression, including the potential attainment of an immune-induced dormant state.
 

[top]