Therapeutic HDAC inhibition in hypermutant diffuse intrinsic pontine glioma

Constitutional mismatch repair deficiency (CMMRD) is a cancer predisposition syndrome associated with the development of hypermutant pediatric high-grade glioma, and confers a poor prognosis. While therapeutic histone deacetylase (HDAC) inhibition of diffuse intrinsic pontine glioma (DIPG) has been reported; here, we use a clinically relevant biopsy-derived hypermutant DIPG model (PBT-24FH) and a CRISPR-Cas9 induced genetic model to evaluate the efficacy of HDAC inhibition against hypermutant DIPG. We screened PBT-24FH cells for sensitivity to a panel of HDAC inhibitors (HDACis) in vitro, identifying two HDACis associated with low nanomolar IC50s, quisinostat (27 nM) and romidepsin (2 nM). In vivo, quisinostat proved more efficacious, inducing near-complete tumor regression in a PBT-24FH flank model. RNA sequencing revealed significant quisinostat-driven changes in gene expression, including upregulation of neural and pro-inflammatory genes. To validate the observed potency of quisinostat in vivo against additional hypermutant DIPG models, we tested quisinostat in genetically-induced mismatch repair (MMR)-deficient DIPG flank tumors, demonstrating that loss of MMR function increases sensitivity to quisinostat in vivo. Here, we establish the preclinical efficacy of quisinostat against hypermutant DIPG, supporting further investigation of epigenetic targeting of hypermutant pediatric cancers with the potential for clinical translation. These findings support further investigation of HDAC inhibitors against pontine high-grade gliomas, beyond only those with histone mutations, as well as against other hypermutant central nervous system tumors.

CBFA2T3-GLIS2 model of pediatric acute megakaryoblastic leukemia identifies FOLR1 as a CAR T cell target

Fusion oncoproteins are the initiating event in the pathogenesis of many pediatric AML. The CBFA2T3-GLIS2 (C/G) fusion is a product of a cryptic translocation primarily seen in infants and early childhood and is associated with dismal outcome. Here, we demonstrate that the expression of the C/G oncogenic fusion protein promotes the transformation of human cord blood hematopoietic stem/progenitor cells (CB HSPCs) in an endothelial cell (EC) co-culture system, that recapitulates the transcriptome, morphology and immunophenotype of C/G AML and induces highly aggressive leukemia in xenograft models. Interrogating the transcriptome of C/G-CB cells and primary C/G AML identified a library of C/G fusion-specific genes that are potential targets for therapy. We developed chimeric antigen receptor (CAR) T cells directed against one of the targets, FOLR1, and demonstrated their pre-clinical efficacy against C/G AML using in vitro and xenograft models. FOLR1 is also expressed in renal and pulmonary epithelium, raising concerns for toxicity that must be addressed for the clinical application of this therapy. Our findings underscore the role of the endothelial niche in promoting leukemic transformation of C/G-transduced CB HSPCs. Furthermore, this work has broad implications for studies of leukemogenesis applicable to a variety of oncogenic fusion-driven pediatric leukemias, providing a robust and tractable model system to characterize the molecular mechanisms of leukemogenesis and identify biomarkers for disease diagnosis and targets for therapy.

Ex silico engineering of cystine-dense peptides yielding a potent bispecific T cell engager

Cystine-dense peptides (CDPs) are a miniprotein class that can drug difficult targets with high affinity and low immunogenicity. Tools for their design, however, are not as developed as those for small-molecule and antibody drugs. CDPs have diverse taxonomic origins, but structural characterization is lacking. Here, we adapted Iterative Threading ASSEmbly Refinement (I-TASSER) and Rosetta protein modeling software for structural prediction of 4298 CDP scaffolds and performed in silico prescreening for CDP binders to targets of interest. Mammalian display screening of a library of docking-enriched, methionine and tyrosine scanned (DEMYS) CDPs against PD-L1 yielded binders from four distinct CDP scaffolds. One was affinity-matured, and cocrystallography yielded a high-affinity (K_D = 202 pM) PD-L1-binding CDP that competes with PD-1 for PD-L1 binding. Its subsequent incorporation into a CD3-binding bispecific T cell engager produced a molecule with pM-range in vitro T cell killing potency and which substantially extends survival in two different xenograft tumor-bearing mouse models. Both in vitro and in vivo, the CDP-incorporating bispecific molecule outperformed a comparator antibody-based molecule. This CDP modeling and DEMYS technique can accelerate CDP therapeutic development.

A potent peptide-steroid conjugate accumulates in cartilage and reverses arthritis without evidence of systemic corticosteroid exposure

On-target, off-tissue toxicity limits the systemic use of drugs that would otherwise reduce symptoms or reverse the damage of arthritic diseases, leaving millions of patients in pain and with limited physical mobility. We identified cystine-dense peptides (CDPs) that rapidly accumulate in cartilage of the knees, ankles, hips, shoulders, and intervertebral discs after systemic administration. These CDPs could be used to concentrate arthritis drugs in joints. A cartilage-accumulating peptide, CDP-11R, reached peak concentration in cartilage within 30 min after administration and remained detectable for more than 4 days. Structural analysis of the peptides by crystallography revealed that the distribution of positive charge may be a distinguishing feature of joint-accumulating CDPs. In addition, quantitative whole-body autoradiography showed that the disulfide-bonded tertiary structure is critical for cartilage accumulation and retention. CDP-11R distributed to joints while carrying a fluorophore imaging agent or one of two different steroid payloads, dexamethasone (dex) and triamcinolone acetonide (TAA). Of the two payloads, the dex conjugate did not advance because the free drug released into circulation was sufficient to cause on-target toxicity. In contrast, the CDP-11R-TAA conjugate alleviated joint inflammation in the rat collagen-induced model of rheumatoid arthritis while avoiding toxicities that occurred with nontargeted steroid treatment at the same molar dose. This conjugate shows promise for clinical development and establishes proof of concept for multijoint targeting of disease-modifying therapeutic payloads.

Optimal therapeutic targeting by HDAC inhibition in biopsy-derived treatment-naïve diffuse midline glioma models

BACKGROUND

Diffuse midline gliomas (DMGs), including diffuse intrinsic pontine gliomas (DIPGs), have a dismal prognosis, with less than 2% surviving 5 years postdiagnosis. The majority of DIPGs and all DMGs harbor mutations altering the epigenetic regulatory histone tail (H3 K27M). Investigations addressing DMG epigenetics have identified a few promising drugs, including the HDAC inhibitor (HDACi) panobinostat. Here, we use clinically relevant DMG models to identify and validate other effective HDACi and their biomarkers of response.

METHODS

HDAC inhibitors were tested across biopsy-derived treatment-naïve in vitro and in vivo DMG models with biologically relevant radiation resistance. RNA sequencing was performed to define and compare drug efficacy and to map predictive biomarkers of response.

RESULTS

Quisinostat and romidepsin showed efficacy with low nanomolar half-maximal inhibitory concentration (IC50) values (~50 and ~5 nM, respectively). Comparative transcriptome analyses across quisinostat, romidepsin, and panobinostat showed a greater degree of shared biological effects between quisinostat and panobinostat, and less overlap with romidepsin. However, some transcriptional changes were consistent across all 3 drugs at similar biologically effective doses, such as overexpression of troponin T1 slow skeletal type (TNNT1) and downregulation of collagen type 20 alpha 1 chain (COL20A1), identifying these as potential vulnerabilities or on-target biomarkers in DMG. Quisinostat and romidepsin significantly (P < 0.0001) inhibited in vivo tumor growth.

CONCLUSIONS

Our data highlight the utility of treatment-naïve biopsy-derived models; establishes quisinostat and romidepsin as effective in vivo; illuminates potential mechanisms and/or biomarkers of DMG cell lethality due to HDAC inhibition; and emphasizes the need for brain tumor-penetrant versions of potentially efficacious agents.

A Protocol for the Generation of Treatment-naïve Biopsy-derived Diffuse Intrinsic Pontine Glioma and Diffuse Midline Glioma Models

Diffuse intrinsic pontine glioma (DIPG) is a universally fatal tumor of the brainstem, most commonly affecting young children. Due to its location, surgical resection is not achievable, but consideration of a biopsy has become standard practice at children’s hospitals with the appropriate neurosurgical expertise. While the decision to obtain a biopsy should be directed by the presence of atypical radiographic features that call the diagnosis of DIPG into question or the requirement of biopsy tissue for clinical trial enrollment, once this precious tissue is available its use for research should be considered. The majority of DIPG and diffuse midline glioma, H3 K27M-mutant (DMG) models are autopsy-derived or genetically-engineered, each of which has limitations for translational studies, so the use of biopsy tissue for laboratory model development provides an opportunity to create unique model systems. Here, we present a detailed laboratory protocol for the generation of treatment-naïve biopsy-derived DIPG/DMG models.

DIPG-10. OPTIMAL HDAC INHIBITION IN DIFFUSE INTRINSIC PONTINE GLIOMA

As the majority of diffuse intrinsic pontine glioma (DIPG) have H3K27M mutations, epigenetic-targeting agents have been studied, though evaluations have been limited by their model systems, untranslatable drug concentrations, and/or evasive mechanisms of action. To develop a more translational model, we used biopsy samples from newly diagnosed DIPG patients to create treatment-naïve in vitro and in vivo models (molecular aberrations in parentheses), including PBT-09FH (H3FA3, PI3KCA), PBT-22FH (H3F3A, TP53), PBT-24FH (PMS2), and PBT-27FH (HIST1H3B, TP53, NTRK2). Models demonstrated radiation-resistance similar to the patient from whom the culture was generated, supporting the models’ relevance (e.g. cell viability after 8 Gy was 36%, 81%, 71%, and 61% in PBT-09FH, -22FH, -24FH, and -27FH, respectively, compared to 7% in the medulloblastoma model MED-411FH). We evaluated cell viability and apoptosis following treatment with a panel of HDAC inhibitors, identifying the low nanomolar IC50 of quisinostat (~50 nM) and romidepsin (~5 nM). While RNA expression changes induced by 100 nM panobinostat and quisinostat included shared overexpression of the top 20/25 genes (e.g. FSTL5, ITIH5) and shared downregulation of the top 22/25 (e.g. GPR37L1, HEPACAM), only 9/25 were downregulated by panobinostat, quisinostat, and romidepsin (e.g. C21orf62, IFIT2), identifying these as potential vulnerabilities or biomarkers of lethal HDAC inhibition. Mass-spectrometry (LC-MS) demonstrated panobinostat as the greatest acetylator of cortactin, potentially related to thrombocytopenia. While PBT-09 flank models demonstrated quisinostat’s on-target acetylation and efficacy, orthotopic xenograft models did not, supporting our model’s intact blood-brain barrier and emphasizing the need for CNS penetrant versions of potentially efficacious agents.

A Protocol for the Generation of Treatment-naïve Biopsy-derived Diffuse Intrinsic Pontine Glioma and Diffuse Midline Glioma Models

Diffuse intrinsic pontine glioma (DIPG) is a universally fatal tumor of the brainstem, most commonly affecting young children. Due to its location, surgical resection is not achievable, but consideration of a biopsy has become standard practice at children's hospitals with the appropriate neurosurgical expertise. While the decision to obtain a biopsy should be directed by the presence of atypical radiographic features that call the diagnosis of DIPG into question or the requirement of biopsy tissue for clinical trial enrollment, once this precious tissue is available its use for research should be considered. The majority of DIPG and diffuse midline glioma, H3 K27M-mutant (DMG) models are autopsy-derived or genetically-engineered, each of which has limitations for translational studies, so the use of biopsy tissue for laboratory model development provides an opportunity to create unique model systems. Here, we present a detailed laboratory protocol for the generation of treatment-naïve biopsy-derived DIPG/DMG models.

A TfR-Binding Cystine-Dense Peptide Promotes Blood-Brain Barrier Penetration of Bioactive Molecules

The impenetrability of the blood-brain barrier (BBB) to most conventional drugs impedes the treatment of central nervous system (CNS) disorders. Interventions for diseases like brain cancer, neurodegeneration, or age-associated inflammatory processes require varied approaches to CNS drug delivery. Cystine-dense peptides (CDPs) have drawn recent interest as drugs or drug-delivery vehicles. Found throughout the phylogenetic tree, often in drug-like roles, their size, stability, and protein interaction capabilities make CDPs an attractive mid-size biologic scaffold to complement conventional antibody-based drugs. Here, we describe the identification, maturation, characterization, and utilization of a CDP that binds to the transferrin receptor (TfR), a native receptor and BBB transporter for the iron chaperone transferrin. We developed variants with varying binding affinities (K_D as low as 216 pM), co-crystallized it with the receptor, and confirmed murine cross-reactivity. It accumulates in the mouse CNS at ~25% of blood levels (CNS blood content is only ~1%-6%) and delivers neurotensin, an otherwise non-BBB-penetrant neuropeptide, at levels capable of modulating CREB signaling in the mouse brain. Our work highlights the utility of CDPs as a diverse, easy-to-screen scaffold family worthy of inclusion in modern drug discovery strategies, demonstrated by the discovery of a candidate CNS drug delivery vehicle ready for further optimization and preclinical development.

TMIC-48. MACROPHAGE DEPLETION COMBINED WITH RADIATION IN A PRECLINICAL MEDULLOBLASTOMA MODEL

BACKGROUND

Radiation induces a measurable immune response, yet this anti-tumor immunity is often suboptimal at eliminating residual brain tumor cells. Medulloblastomas, along with other pediatric brain tumors, tend to recur in the primary resection bed following irradiation. Infiltrating macrophage within the tumor (TAM) are radioresistant immune cells that may limit the effectiveness of radiotherapy and promote tumor regrowth through their immunosuppressive and pro-tumorigenic functions. Therefore, we explored whether CSF-1 receptor (CSF1R) inhibition, and essentially TAM depletion, could augment radiation and/or reduce local recurrence.

METHODS

Selected CSF1R inhibitors were tested on both human and mouse macrophage in vitro. In vivo studies were performed on a patient-derived orthotopic xenograft (PDOX) model of Group 3 medulloblastoma, Med411-FH. Macrophage depletion was achieved using the anti-murine CSF1R antibody, AFS98, injected intra-peritoneally 10 days prior to, concurrently, and after radiation.

RESULTS

CSF1R antibodies and small molecule tyrosine kinase inhibitors were effective at stanching macrophage survival in culture. In a PDOX medulloblastoma model, both classes of CSF1R inhibitors were effective at reducing macrophage in the brain after approximately 10 days of administration, with maximal effect of 90% depletion achieved after 21 days. In a combinatorial study of AFS98 and irradiation, TAM depletion conferred a very modest effect on tumor growth (p=0.003) and had no effect on overall survival when combined with radiation (p=0.86). Interestingly, we did observe a trend toward reduction in metastatic spinal disease following radiation in the AFS98-treated cohort.

CONCLUSION

CSF1R antibody administration effectively eliminated macrophage and microglia in the brain and spine, suggesting adequate blood-brain barrier penetration in our xenograft model. While CSF1R inhibition combined with radiation was ineffective at prolonging survival, further studies are needed to investigate a modest effect on brain tumor growth and an observed (yet not statistically significant) difference in metastatic spinal recurrence in drug-treated animals.

Calibration of fluorescence imaging for tumor surgical margin delineation: multistep registration of fluorescence and histological images

Although a greater extent of tumor resection is important for patients' survival, complete tumor removal, especially tumor margins, remains challenging due to the lack of sensitivity and specificity of current surgical guidance techniques at the margins. Intraoperative fluorescence imaging with targeted fluorophores is promising for tumor margin delineation. To verify the tumor margins detected by the fluorescence images, it is necessary to register fluorescence with histological images, which provide the ground truth for tumor regions. However, current registration methods compare fluorescence images to a single-layer histological slide, which is selected subjectively and represents a single plane of the three-dimensional tumor. A multistep pipeline is established to correlate fluorescence images to stacked histological images, including fluorescence calibration and multistep registration. Multiple histological slices are integrated as a two-dimensional (2-D) tumor map using optical attenuation model and average intensity projection. A BLZ-100-labeled medulloblastoma mouse model is used to test the whole framework. On average, the synthesized 2-D tumor map outperforms the selected best slide as ground truth [Dice similarity coefficient (DSC): 0.582 versus 0.398, with significant differences; mean area under the curve (AUC) of the receiver operating characteristic curve: 88% versus 85.5%] and the randomly selected slide as ground truth (DSC: 0.582 versus 0.396 with significant differences; mean AUC: 88% versus 84.1% with significant differences), which indicates our pipeline is reliable and can be applied to investigate targeted fluorescence probes in tumor margin detection. Following this proposed pipeline, BLZ-100 shows enhancement in both tumor cores and tumor margins (mean target-to-background ratio: 8.64 ± 5.76 and 4.82 ± 2.79 , respectively).

A new approach for immuno-oncology biomarker discovery: High-plex, spatial protein profiling based on NanoString digital quantification

27

Background: The immune response to cancer is shaped by the abundance and localization of immune cell populations, their activation status, and expression of immunomodulatory factors. To detect proteins at high multiplex with spatial resolution, NanoString optical barcoding technology was used to digitally profile protein expression in formalin fixed paraffin embedded (FFPE) samples and compared to traditional IHC. Methods: NanoString has enabled digital profiling of immuno-oncology protein targets, including immune cell markers and checkpoint proteins. Protein detection is enabled via primary antibodies (Abs) which are attached via a UV cleavable linker to DNA indexing oligos. FFPE samples are stained with a multiplex cocktail of labeled Abs, then DNA oligos are released by UV light exposure. Liberated oligos are then hybridized to optical barcodes for quantitation on a NanoString instrument. This technique enables quantitative, multiplex protein detection over 5 logs of dynamic range. IHC and NanoString spatial protein profiling were performed on alternating sections from blocks of tonsil, melanoma, and colorectal cancer. Sections were fluorescently labeled with panCK, Ki67, and Syto-83 to visualize morphology, and regions of interest (ROI) were selected for profiling using up to 30 oligo-tagged Abs. Results: NanoString counts strongly correlated with quantification of CD3, CD4, CD8, PanCK, Ki67, PD-1, and PD-L1 derived from IHC. To explore limits of detection, ROIs of 1, 2, 4, or 8 cells were profiled. Digital counts were detected above background at the single cell level and linearly correlated with cell count. Furthermore, quantification with two cell lineage markers and double positive cells were demonstrated using a 30-plex Ab cocktail. Conclusions: Multiplex, digital protein profiling with spatial resolution will enable deep characterization of immune responses in tumors. Additionally, single cell profiling enables inter-cellular characterization of variations in immune response. The strong correlation of NanoString data to IHC indicates the feasibility to spatially profile multiple key proteins with minimal consumption of patient tissue.

Abstract A66: Runx3 drives metastatic programs in pancreas cancer

The extreme proclivity of pancreatic ductal adenocarcinoma (PDA) for metastatic spread is the principal cause of mortality for this disease, even for patients who successfully undergo surgical resection of the primary tumor. We have engaged in a concerted effort to define the fundamental molecular and cellular drivers of PDA through the use of cognate genetically engineered mouse models that recapitulate the spectrum of human PDA. Recently, we incorporated conditional deletion of Smad4 into the KPC mouse model and noted a significant attenuation of metastatic burden in the resultant KPDC mice. Gene expression analysis identified the Runt family transcription factor Runx3 as highly expressed in metastatic KPC tumors, but minimally expressed in PDA from KPDC mice. We showed that aberrant expression of Runx3 strongly enhanced the metastatic potential of PDA through an orchestrated program that stimulated cancer cell invasion and migration and also helped generate a niche for disseminated cells to establish themselves in foreign tissues (such as the liver or lung) and form metastases. Runx3 expression additionally inhibited proliferation by upregulating expression of the cell cycle inhibitor Cdkn1a (p21), initiating a metastatic switch that favored PDA dissemination over local outgrowth. We have also identified influences of Runx3 expression on key developmental and immune-regulatory signaling pathways that are known to be aberrantly regulated in PDA. Because Runx3 has been shown to interact with Smad4 and p53, the genomic context of PDA expressing Runx3 critically modulates its function through de novo or altered protein complex formation. Defining the dichotomous, contextual roles of Runx3 as both a tumor “suppressor” and “promoter” may provide insight into the proximal cause of lethality in patients, thereby informing clinical prioritization of local versus systemic therapies for individual patients with PDA.

Citation Format: Martin C. Whittle, Libing Feng, Ashley Dotson, Fiona Pakiam, Sunil R. Hingorani.{Authors}. Runx3 drives metastatic programs in pancreas cancer. [abstract]. In: Proceedings of the AACR Special Conference on Pancreatic Cancer: Advances in Science and Clinical Care; 2016 May 12-15; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2016;76(24 Suppl):Abstract nr A66.