TARGETING SURVIVIN IN CANCER

Abstract

With almost 4,000 citations in Medline in a little over ten years, survivin has certainly kept scores of investigators busy worldwide. Tangible progress has been made in revealing the multiple functions of survivin, uncovering their wirings as integrated cellular networks, and mapping their exploitation in virtually every human tumor, in vivo. Considering the normally long and excruciating timeline of oncology drug discovery, it is clearly a resounding success that a better understanding of survivin biology has led to several clinical trials of survivin-based therapeutics in cancer patients. However, the portfolio of survivin antagonists available in the clinic remains small, pressing the need for a less rigid drug development approach to fully unlock the potential of this unique, albeit unconventional oncology drug target.

1. Overview of survivin

This contribution is not designed to provide a full review on survivin. Comprehensive articles on these topics have been published, which may provide readers with a more in-depth perspective [1,2,3,4,5,6,7]. However, a brief summary of the biology of survivin is presented below to help frame the challenges and opportunities associated with survivin-based therapeutics in cancer.

Survivin is the smallest member of the Inhibitor of Apoptosis (IAP) gene family [8], containing a single Baculovirus IAP Repeat (BIR), which is the hallmark of these molecules. Although a homodimer by X-ray crystallography [8], survivin can function as a monomer for at least some protein-protein interactions, as well as mechanisms of subcellular localization [9]. The control of survivin gene expression at the tip of chromosome 17 in humans (17q25) is complex, and involves scores of positive and negative regulators. Several oncogenic transcription factors stimulate expression of the survivin gene, whereas multiple tumor suppressors actively repress the survivin gene [10]. The dynamic interplay in this transcriptional network controls a unique feature of survivin, namely its sharp differential expression in cancer, as opposed to normal tissues, in vivo [6,7]. Transcription of the survivin gene is strictly cell cycle-regulated, and peaks at mitosis [1]. However, examples of cell cycle-independent survivin gene expression have been described, especially in response to growth factor or cytokine stimulation [6]. A survivin protein is extensively post-translationally modified by degradative and non-degradative cycles of ubiquitylation and de-ubiquitylation, as well as phosphorylation [6], which control protein stability, binding to molecular partners and trafficking to various subcellular compartments.

Survivin is a multifunctional protein and is essential, in that constitutive or conditional deletion of the survivin gene is incompatible with tissue or organism viability [6]. Orthologs of survivin have been found in lower organisms, such as yeast, worms, and flies, suggesting evolutionary conservation of this pathway. In mammalian cells, survivin participates in at least three homeostatic networks: the control of mitosis (1), the regulation of apoptosis (2), and the cellular stress response (3). This classification is not restrictive, as novel functions of surviving are continuously proposed, as well as new roles for known properties. Even within the same network, survivin plays multiple roles. For instance, at mitosis, survivin acts as a passenger protein [11] for proper chromosomal alignment, controls chromatin-associated spindle formation [12], enhances spindle stability via suppression of microtubule dynamics [13], and oversees kinetochore–microtubule attachment [14] in the spindle assembly checkpoint.

As a cytoprotective molecule [6,7], survivin, like all other IAPs except XIAP [8], does not directly inhibit caspases, but interacts with protein partners, most notably XIAP [15]. This complex promotes increased XIAP stability against degradation, activates multiple signaling pathways, including NFκB, synergistically inhibits caspase-3 and -9, suppresses apoptosis, and accelerates tumor progression, in vivo. Other mechanisms of survivin cytoprotection have been proposed, including the ability of mitochondrial-localized pool of survivin to sequester pro-apoptotic Smac away from XIAP [16], or altogether prevent its release from mitochondria [17].

A third function of survivin in the cellular stress response is beginning to emerge, and involves the association of survivin with various molecular chaperones, including the aryl hydrocarbon receptor-interacting protein (AIP) [18], Hsp60 [19], and Hsp90 [20]. These interactions may promote adaptation under conditions of cellular stress by maintaining survivin protein stability, folding, and subcellular trafficking, in vivo.

2. Targeting survivin in cancer

Because of its nodal properties, and over-expression in virtually every human tumor, survivin has been intensely pursued as a drug target for novel cancer therapeutics [5,6,7]. Consistent with its multifunctional roles, the presence of survivin in cancer has been universally linked to resistance to apoptosis, metastasis, bypass of cell cycle checkpoints, and resistance to therapy. This has been further bolstered by molecular profiling studies and retrospective analyses of patient cohorts, which have consistently identified survivin as a ‘risk factor’ for poor prognosis and disease recurrence [21]. Against this backdrop, putative survivin antagonists would not only function as single-gene inhibitors, but act more broadly as ‘pathway antagonists’ [22], disabling the multiple survivin networks, and thus ideally suited to overcome tumor heterogeneity, in vivo.

2.1 - Molecular antagonists

The first described molecular antagonist of survivin is a phosphorothioate antisense oligonucleotide [23], which suppressed survivin mRNA and protein expression, and produced strong anticancer activity in preclinical models. Sponsored by Ely Lilly and Co., this agent, designated LY2181308 showed a favorable toxicity profile with evidence of survivin downregulation in a phase I trial in patients with advanced cancers [24]. A locked nucleic acid antisense oligonucleotide to survivin has also been developed by Santaris Pharma, and, more recently, ENZON Pharmaceuticals [25]. A phase I clinical trial with this agent, designated SPC3042 (or ENZ3042), also showed excellent safety [26]. Additional strategies to acutely lower survivin levels in tumor cells involved small interfering RNA (siRNA) or hammerhead ribozymes [27]. In preclinical studies, these agents consistently showed anticancer activity, alone or in combination with chemotherapy, with no detectable systemic toxicity [27]. The delivery of siRNA in vivo is challenging, but the apparent success of recent studies [28] suggests that survivin-directed gene silencing may at some point be evaluable in cancer patients.

2.2 - Small molecules

Small molecules that directly target survivin have been developed [6], and several clinical trials with these agents have been completed, with more underway. The most extensively studied agent in this class is YM155 [29]. Originally developed by Yamanouchi Pharmaceuticals, and, more recently, Astellas Pharma, YM155 was identified in high-throughput screening for inhibitors of survivin promoter activity [29]. This effect was specific for the survivin gene, and resulted in potent anticancer activity in preclinical models [29]. Two phase I studies of YM155 in heavily pretreated cancer patients have been published. The trial conducted in the US reported impressive responses, with tumor shrinkage and durable remissions in patients with advanced prostate cancer, large cell non-Hodgkin’s lymphoma and non-small cell lung cancer [30]. The Japan phase I trial of YM155 also provided evidence of disease stabilization in nine patients [31]. Importantly, both studies showed a favorable toxicity profile with minimal and rapidly reversible side effects. Two phase II studies of YM155 monotherapy have been recently published. In patients with advanced non small cell lung cancer, YM155 produced two partial responses and disease stabilization in 14 out of 37 evaluable patients, corresponding to a disease control rate of 43% [32]. Consistent with the phase I data, the treatment was well tolerated with the majority of treatment discontinuations not treatment related [32]. The results of a phase II study of YM155 in melanoma were less encouraging. Although well tolerated, YM155 produced only one partial response in patients with stage III and stage IV disease, failing to meet the pre-specified endpoint of two responses in 29 evaluable patients [33].

Another direct small molecule inhibitor of survivin is tetra-O-methyl nordihydroguaiaretic acid (M(4)N), which also acts as a transcriptional repressor of the survivin promoter, potentially by antagonizing Sp1-dependent gene expression [34]. This compound, designated Terameprecol (EM-1421) [35] has shown good preclinical activity with an impressive 88% bioavailability, in vivo [36]. Terameprecol has been formulated for systemic delivery to cancer patients, and a phase I study in patients with advanced solid malignancies has shown favorable safety and disease stabilization in 8 out of 25 evaluable patients [35]. Another phase I study of Terapremecol in 16 heavily pretreated patients with adult myelogenous leukemia (AML) has also shown favorable safety, one partial response and disease stabilization in 5 patients [37]. In addition, Terameprecol has been formulated as a 1% or 2% vaginal ointment for local application in women with papillomavirus- or herpes simplex virus-associated carcinogenesis. Two phase I studies with Terameprecol ointment have shown excellent safety, no adverse events and no systemic absorption of the agent [38,39].

2.3 - Cancer vaccine/immunotherapy

Because of its differential expression in cancer, as opposed to normal tissues, it has been hypothesized that cancer patients may recognize survivin as a non-self protein, and mount an immune response to it [40]. This concept has been validated in the clinic, and sera from cancer patients contain antibodies [41], and cytolytic T cells against survivin [42]. This immune recognition has been mapped in detail [43,44], and dendritic cells pulsed with survivin peptides, survivin-containing tumor lysates or transduced/transfected with survivin, elicit cytolytic T cell responses and MHC-restricted anticancer activity in vitro [45,46], and in preclinical models [47]. Several phase I/II trials of survivin-directed immunotherapy have been reported [40,48]. In these studies, survivin-based vaccination was invariably safe, devoid of significant side-effects, and associated with antigen-specific immunologic responses. Although no objective responses were noted, two phase I/II trials with infusion of dendritic cells pulsed with survivin showed durable (>6 months) disease stabilization in 24% of melanoma patients [49], and 50% of renal cell carcinoma patients [50]. A phase I/II trial of systemic delivery of protamine-protected survivin mRNA in melanoma was also safe, produced detectable T cell responses, and achieved one complete response out of seven evaluable patients [51]. Anecdotal reports also suggest that survivin-based vaccination may be effective against metastatic disease [52].

3. Survivin-based therapeutics: what’s ahead?

Given the herculean task of oncology drug discovery and development [53], moving from discovery to the clinic in ten years is undoubtedly a resounding success for the survivin field. The approaches described above remain promising, and their excellent safety profile bodes well for their continued evaluation in the clinic. There were concerns that the low but detectable level of survivin in normal tissues [54], and its essential role in stem cells [55], could compromise the exploitation of this target, in vivo. In reality, this has not occurred, which may reflect a mechanism of ‘survivin addiction’ [56] specifically by the tumor, or, alternatively, the existence of ‘qualitative’ differences in the wiring of survivin networks in transformed cells, as opposed to normal tissues [6].

For the next wave of survivin-based therapeutics, combination studies with cytotoxics or radiation appear as obvious choices. Hopefully, these protocols will take advantage of the better understanding of survivin biology accumulated over the past decade. For instance, sequential inhibition of survivin in tumor cells pretreated with taxanes, and thus arrested at mitosis with hyperpolarized microtubules, produces a much increased anticancer activity with no toxicity in preclinical models [57], so that a similar sequential combination strategy could be envisioned for survivin small molecule or antisense antagonists. Conversely, suppression of survivin has been invariably shown to sensitize disparate tumor cells to ionizing radiation, and this mechanism can be also readily exploited for novel radiotherapy protocols.

There may be also concrete opportunities to rationally combine survivin antagonists with various molecular agents. For instance, survivin is a target gene of various oncogenic signaling circuits, including STAT3, Akt, EGFR [6], Notch [58], and Wnt [59]. Although small molecule inhibitors of these pathways have produced modest gains in the clinic as monotherapy, it is possible that the combined inhibition of one of their critical downstream targets, i.e. surviving may enhance and broaden their efficacy, without compromising safety. Finally, because surviving is a target of both VEGF [60] and Ang-1 [61], and is critically required for endothelial cell viability [62], its inhibitors could be rationally incorporated in current anti-angiogenic regimens, for instance in patients with renal cell cancer or hepatocellular carcinoma receiving multi-kinase inhibitors.

Despite these hopeful opportunities, the reality is that the current portfolio of surviving antagonists remains small, and unlikely to fully unlock the potential of this target for novel cancer therapeutics. The multiple and likely redundant regulation of the survivin promoter in tumor cells [10] poses formidable challenges for agents like YM155 or Terameprecol to durably shut off survivin gene expression in disparate and highly heterogeneous cellular contexts. Antisense molecules have been invariably plagued by issues of specificity, off-target effects or lack of efficacy, and siRNA-based therapeutics is still in its infancy, having to overcome fundamental hurdles in delivery, durable gene silencing, and, ultimately, long-term efficacy in a highly heterogeneous disease.

The lack of a larger portfolio of survivin antagonists likely reflects a parsimonious view that molecules that are not present on the cell surface, or lack inhibitable catalytic activity, are not good targets [22]. For survivin, crystallographic data reveal little in the way of potential drugable sites, typically structural ‘pockets’ of suitable geometry and hydrophobicity. Despite these challenges, the wealth of information collected in the past decade about survivin should open fresh opportunities for drug discovery. For instance, disruption of protein–protein interactions, especially those involving apoptosis regulators, is now feasible, and has generated agents with promising anticancer activity [63]. In this context, proof-of-principle agonists that disrupt the complex between survivin and other network components — for instance, Hsp90 — have been identified [64], and validated for anticancer activity with no toxicity for normal tissues in preclinical studies [64]. Accordingly, it may be possible to envision high-throughput screening strategies to identify small molecule inhibitors of the interaction between survivin and some of its critical partners, for instance Smac [65], XIAP [15], or members of the chromosomal passenger complex [9], which may generate agents with novel specificities and anticancer efficacy. What we have learned over the past decade is that even seemingly small perturbations in survivin protein turnover, post-translational modifications, subcellular localization or binding to molecular partners, has almost invariably catastrophic consequences for tumor cells, resulting in mitotic defects, cell cycle arrest, spontaneous apoptosis, loss of metastatic potential, and inhibition of tumor growth, in vivo. As for the survivin antagonists now in the clinic, such novel agents can also be expected to have broad activity in genetically disparate tumor types, a favorable toxicity profile, and synergistic efficacy with conventional or targeted anticancer regimens.