One of the most critical steps for effectively treating GBM is to identify pathways selectively altered in tumor cells that can be exploited by targeted therapies. As such, a central component of our work has been to determine the expression levels of various tumor-specific receptors (death receptors, EGFR, IL13Ra2, etc) in primary and established GBM lines, and subsequently develop novel therapies targeting these receptors. We utilize non-invasive imaging to monitor both changes in the targets (receptors) and the therapeutic efficacy of newly developed therapies at multiple resolutions.

In a study reported in Cancer Research, we investigated the role of the epidermal growth factor receptor (EGFR) in GBM progression and it’s response to targeted therapies, as EGFR is upregulated or mutated in up to 60% of GBM cases. We first engineered GBM lines with diagnostic variants of two different EGFR receptors. Using real-time quantitative imaging of receptor levels and GBM volumes (Figure 1) we showed there is a direct correlation between EGFR expression and GBM progression, and demonstrated that down-regulation of EGFR by targeted therapies reduces GBM burden in vivo. These results demonstrate the use of bioimaging to monitor multiple events in the response of GBM, and show the effectiveness of developing therapies to target GBM-specific receptors. In ongoing work, we have engineered a variety of therapies targeted to different GBM-specific receptors (see below) as well as additional fusions between tumor-specific targets and optical reporters for diagnostic tracking. We have also utilized vectors encoding receptor-driven luciferase reporters that allow us to track endogenous receptor levels and the up-regulation of these tumor-specific targets by various molecules in real-time in vivo using serial bioimaging.

Figure 1. (A-D) GBM volumes and EGFR expression visualized by quantitative bioluminescence imaging (A-B), high-resolution intravital microscopy (C), or post-mortem IHC (D) performed on mice implanted with human Gli36-Fluc-DsRed2/EGFR-GFP-RLuc cells. (E) Quantitative BLI showing expression of a diagnostic fusion of the GBM-specific receptor IL13R2. (F) Bioluminescence imaging performed on mice implanted with GBM expressing death receptor luciferase reporters

Genetically modified neural stem cells hold enormous promise for treating a variety of neurological diseases due to their ability to give rise to cells of neural lineage. More importantly for our studies, work by our group and others has demonstrated these cells also possess a unique capacity to home to diseased areas of the brain, including GBM deposits, where they can provide robust and on-site delivery of therapeutics. Therefore, stem cells represent a powerful vehicle capable of delivering therapies to localized and diffuse GBM cells that are not typically accessible with standard methods of therapeutic administration (i.e.: systemic chemotherapies, intracranial wafers, intraparenchymal viral injection). Further, they also overcome the limitations of short circulating half-life and non-specific toxicities hindering traditional methods of delivery.

However, easy serial monitoring of stem cells and GBM volumes in vivo is vital to the success of this approach. Therefore, we have developed and employed a library of viral vectors (lentivirus, retrovirus, adeno-associated virus; partial list included in Figure 2) that express: 1) combinations of fluorescent and bioluminescent fusion proteins to stably modify cells for diagnostic tracking or 2) receptor-targeted anti-tumor therapies (see area #3). The application of this technology was highlighted in a recent publication (J. Neurosci, 2008) where we employed high resolution intravital microscopy to track the intracranial migration of human neural stem cells (hNSC) towards GBM deposits in real-time (Figure 2). Simultaneously, quantitative dual bioluminescence imaging showed both the immune system and the presence of tumors had significant influences on the survival of hNSC. Of equal importance, these results demonstrate our ability to non-invasively quantify and track stem cell volumes and intracranial GBM burden, parameters that are vital to our studies investigating the therapeutic efficacy of stem cell-delivered anti-tumor therapies.

Building on this foundation of LV-encoded diagnostic vectors and multi-modality imaging, we have markedly expanded our library of diagnostic vectors, therapeutic vectors (see area #3), available stem cell lines, and GBM cell lines (see area #4). Currently, we work with several human and mouse neural stem cell lines, human and mouse mesenchymal stem cells, mouse embryonic stem cells, and mouse iPS cells. Our current and future interests include characterizing the value of these various stem cell types as therapeutic delivery vehicles in a variety of different patient-derived GBM models (see below for details). A central component of this work will be the continued integration of quantitative non-invasive bioimaging (optical and PET; see below) to elucidate various events in stem cell-mediated therapeutic delivery.

Figure 2. (A) Examples of diagnostic and therapeutic viral vectors expressing fluorescent and bioluminescent fusions, the anti-tumor therapeutic TRAIL, a modified variant of the anti-cancer cytokine mda-7 (SM7L, see below), and a diagnostic variant of EGFR. (B-E) Non-invasive imaging showing the migration of Fluc-mCherry labeled hNSC (B,D) towards GFP-Rluc labeled gliomas (C,E) using high-resolution intravital microscopy (B-C) or dual bioluminescence imaging (D-E).

Due to the current lack of effective treatments for GBM, developing new and more effective anti-tumor molecules and delivery methods for this highly aggressive malignancy is crucial to improving the survival of patients suffering GBM. Utilizing the advantages of stem cell-based delivery, our past and current research has focused on the engineering and validation of several receptor-targeted molecules optimized for cell-based delivery. In a recent study published in the journal Stem Cells, we reported the engineering and application of SRL_OL₂TR, a unique multifunctional fusion variant of the potent pro-apoptotic protein TNFa-related apoptosis-inducing ligand (TRAIL) (Figure 3), to investigate multiple parameters in stem cell-based drug delivery. SRL_OL₂TR allowed the first real-time non-invasive tracking of stem cell-delivered therapies by incorporating diagnostic and therapeutic properties into the molecule. Utilizing SRL_OL₂TR and quantitative bioimaging to provide simultaneous feedback on stem cell volumes, TRAIL therapy, and tumor volumes in vivo, we showed: A) cell-based delivery improves pharmacokinetics and bio-distribution of therapies that significantly increases their anti-GBM efficacy in vivo (Figure 3), and B) variations in the secretion rates and survival of different stem cell lines effects their therapeutic potential.

Utilizing this platform, we are continuing to develop new and novel targeted therapies. We are exploring their delivery from various stem cell lines as well as exploring their utility in combination with clinically relevant therapies that include Temozolomide chemotherpay and radiation.

Figure 3. Peripheral GBMs were treated with either stem cell-delivered SRLOL2TR (A-C) or an intravenous bolus of media containing SRLOL2TR (D-F). Various parameters of SRLOL2TR delivery and efficacy were followed by non-invasive imaging. (A-C) Images and summary data demonstrating delivery of SRLOL2TR by engineered stem cells was stable for greater than 24 hrs (A), reduced the volume of established GBM by day 2 (B), and was localized predominantly at the tumor when ex vivo analysis of excised tissue was performed (C). (D-E) Images and summary data showing SRLOL2TR delivered by intravenous injection was rapidly cleared by 24 hrs (D), was widely distributed in various tissues types (E), and had no effect on GBM progression (F).

The ultimate goal of my research is to develop new therapies that can be successfully translated to clinics to improve the survival of patients suffering from GBM. Therefore, preclinical models that more accurately mimic the clinical scenarios of GBM therapy and new approaches to translate stem cell-based therapies into clinics are crucial for maximizing the predictive potential of preclinical studies. To this end, we have played a central role in developing a new preclinical model of GBM resection and recurrence that integrates primary patient-derived GBM lines in order to address several vital issues including: A) Current preclinical models of GBM rely on treating solid tumors and do not reflect the clinical scenario of GBM resection, and B) Establishing methods that allow retention of therapeutic stem cells in the cavity following GBM surgical de-bulking. In a very recent study (Nat. Neurosci., 2011), we used multi-modality imaging to resect human GBM in the mouse brain (Figure 4). We then seeded therapeutically engineered stem cells encapsulated in biodegradable synthetic extracellular matrix into the resection cavity. This approach prevented cell “washout” from the resection cavity, a significant hurdle to the clinical translation of cell-based therapies for GBM. Retention of encapsulated cells in the cavity lead to robust longitudinal secretion of therapies from the stem cells that significantly delayed tumor recurrence and increased survival. We are now using this highly predictive model to investigate the effectiveness of new therapies against a panel of more clinically relevant patient derived GBM-initiating lines with varying degrees of invasion (high, medium, and low) as these lines more accurately recapitulate human GBM. We are also investigating the therapeutic benefits of delivering additional “armed” stem cell lines as well as viral vectors encapsulated and seeded in resection cavities.

Figure 4. (A) Photomicrograph showing the cavity (*) after resection of human GBM tumor in mice. (B-D) Light images (B) or fluorescent intravital microscopy images (C-D) showing stem cells encapsulated in synthetic extracellular matrix and seeded into the resection cavity. Arrows indicate residual mCherry-labeled tumor cells. (E) Summary data showing the survival of mice following tumor resection and treatment with encapsulated stem cells engineered with S-TRAIL or control (GFP). (F-H) Fluorescent images showing tumor formation by GBM4 (low invasion, F), BT74 (medium invasion, G), and GBM8 (highly invasive, H) primary patient-derived human GBM lines transduced with LV-mCherryFLuc.