Now in its third year, the Selective Targeting of Pancreatic Cancer (SToP) Specialized Program of Research Excellence (SPORE) grant has teamed up with UNC Lineberger’s Developmental Research Program (DRP) to support novel projects seeking to gain key insight into critical areas of pancreatic cancer research. During each year of this five-year grant, we have the unique opportunity to fund the work of junior and established investigators whose work is directed toward elucidating mechanisms of pancreatic cancer development, progression or therapy.
Here is an overview of the innovative research we’ve been able to support through the SToP Cancer SPORE DRP in its first two years, including each of the project’s goals and the investigators’ significant findings and conclusions.
Kirsten Bryant, PhD, and John Morris, PhD: “Identifying Determinants of PDAC Heterogeneity that Predict Response to Dual ERK MAPK and Autophagy Inhibition”
Bryant and Morris sought to leverage mouse models of pancreatic ductal adenocarcinoma (PDAC) to identify determinants of initial response and resistance to concurrent inhibition of the RAS and autophagy pathways to facilitate the clinical advancement of this therapeutic strategy.
Their first aim was to determine the functional determinants of stage-dependent response to concurrent RAS and autophagy inhibition. Their preliminary data focused on a panel of organoids that reflect the distinct stages of PDAC evolution following p53 loss. First, copy number acquisition is observed in diploid cells that frequently display pre-malignant morphology (“early” clones), followed by extensive copy number rearrangement and polyploidy in frank PDAC (“late” clones). They demonstrated that early LOH clones were more sensitive to both MEK and autophagy inhibition.
Furthermore, they applied reverse phase protein array (RPPA) as an orthogonal approach to understand the consequences of both MEK and autophagy inhibition alone and in combination across their panel of evolutionarily staged organoid models. They identified stage-dependent differences in response to treatment with MEKi and CQ alone and in combination, particularly within the ERK MAPK pathway and cell cycle regulatory proteins. They plan to extend these studies by using phosphoproteomic profiling to identify how global stage dependent differences evolve during PDAC development.
The second aim of their project was to delineate determinants of intra- and inter-tumoral heterogeneity in autophagic adaptation and pre-clinical response and resistance to concurrent inhibition of the RAS pathway and autophagy. Progress toward this aim has been focused on efforts to elucidate a more potent and specific autophagy inhibition strategy. To identify more effective autophagy targets, the Bryant lab performed a CRISPR-Cas9 mediated genetic loss-of-function screen in PDAC cell lines and found that loss of the PIKFYVE gene (which encodes the lipid kinase PIKfyve) significantly impaired proliferation of KRAS-mutant PDAC cell lines. PIKfyve inhibition resulted in a robust vacuolarization of the cytoplasm. These vacuoles are easily visible by microscopy in 2D culture systems, and their appearance correlates with the GI50 of each individual cell line.
The Morris lab then evaluated whether PIKfyve inhibition with apilimod could induce the formation of vacuoles in vivo. Using an orthotopic KPC-derived model treated with either vehicle or apilimod, they performed H&E staining and revealed profound vacuole formation in PIKfyvei-treated tumors relative to vehicle control. Consistent with on-target inhibition of PIKfyve, vacuole accumulation in tumors corresponded with accumulation of the autophagosome marker LC3B-II. Additionally, vacuolized tumor regions displayed notably reduced Ki67 staining, demonstrating the antiproliferative effects of PIKfyvei treatment in vivo. Ultimately, Bryant and Morris concluded that PIKfyve inhibition could represent a more specific and potent anti-autophagy target. These findings were recently published in Cancer Research and the data resulting from this development award has supported a Co-PI R21 grant submitted by Drs Bryant and Morris scheduled for review in June 2025.
Clint Stalnecker, PhD: “Investigating the molecular consequences of direct KRAS inhibition in pancreatic cancer”
To determine the molecular consequences of direct KRAS inhibition in pancreatic cancer, Stalnecker treated a panel of pancreatic cancer cell lines with inhibitors selective for KRAS or the downstream KRAS effector kinase, ERK (ERKi), and performed comprehensive phosphoproteomics and proteomics. He evaluated two distinct KRAS inhibitors—a covalent KRASG12D-selective inhibitor (G12Di) or a pan-RAS inhibitor (RASi)—to determine if mutant-selective inhibition would result in distinct rebound activities compared to inhibition of all RAS proteins or downstream signaling through ERK. Stalnecker treated cells for 72 hours to evaluate compensatory rebound activities, and performed phospho-tyrosine, phospho-serine/threonine, and global proteomics. By 72 hours, Stalnecker found nearly identical regulation of downstream signaling between ERK, KRASG12D-selective, and pan-RAS inhibitors.
These observations are consistent with Stalnecker’s previous studies that identified ERK as the principal downstream effector for KRAS. Surprisingly, few differences were observed between G12Di, RASi, and ERKi, except at the level of ERK itself. He found upregulation of ERK phosphorylation in G12Di and ERKi, but not RASi. These results are consistent with rebound activities reported for mutant-selective KRAS inhibitors.
Stalnecker found that 72-hour treatment of cells with G12Di, RASi, and ERKi caused a dramatic upregulation of kinases, and in particular kinases related to RHO GTPase signaling. He combined this dataset with previous data from pancreatic cancer cell lines treated acutely with 1-hour ERKi or 24-hour ERKi and found dynamic kinome reprogramming. Combining these datasets revealed a temporal inhibition of cyclin-dependent kinase activities at 24 hours that began rebounding by 72 hours. Additionally, RHO effector kinases were significantly upregulated at 72 hours, but not at either earlier time points. These results show signaling adaptation to KRAS/ERK inhibition that involves RHO GTPase signaling.
Stalnecker’s observations are significant because it is well known that resistance limits the durability of anti-RAS therapies, and therefore combination approaches will be required to improve the clinical benefit of RAS inhibition. It is suspected that signaling adaptation is a primary response that can lead to resistance. Stalnecker was co-author on a paper that recently reported activation of the HIPPO pathway components YAP1/TAZ as a mechanism of both innate and acquired resistance to KRAS inhibition (Edwards et al. 2023 Can. Res., Wasko et al. 2024 Nature). YAP1/TAZ activation has been causally linked to RHO GTPase-dependent signaling, and Stalnecker’s phosphoproteomics dataset identified upregulation of RHO GTPase effector kinases. His results identified upregulated RHO-associated protein kinases (ROCK1/2) and p21 activated kinases (PAK1/2/3). This led him to evaluate ROCK inhibition (ROCKi) or PAK inhibition (PAKi) in cells with acquired resistance to RASi, where he found RASi resistant cells were more sensitive to ROCKi or PAKi compared to their parental counterparts.
Collectively, these observations support a model whereby RAS or ERK inhibition leads to increased RHO GTPase activity that stimulates ROCK and PAK. Stalnecker hypothesizes that this increased RHO GTPase activity can regulate YAP1/TAZ to drive resistance. To test this, he treated cells resistant to RASi with ROCKi or PAKi and looked for decreased YAP1/TAZ output. He found inhibiting ROCK or PAK decreased expression of the YAP1/TAZ target gene CYR61 in pancreatic cancer cells that have developed resistance to RAS inhibition. Notably, he also stronger growth inhibition and observed an increase in the apoptotic marker cleaved PARP, suggesting this therapeutic strategy induces cell death in cells resistant to RASi.
Stalnecker concluded that sustained inhibition of RAS, KRAS G12D, or ERK led to global reprogramming of the proteome and phosphoproteome. He found targeting RAS, mutant-selective KRAS G12D, or the downstream effector ERK resulted in nearly identical changes in the proteome and phosphoproteome. The phosphoproteomics identified upregulation of many kinases, including the RHO GTPase effector kinases ROCK and PAK, and inhibiting these kinases could sensitize cells to RAS inhibition.
These studies provided the necessary preliminary data to support Stalnecker’s R01 grant proposal submitted to the NCI, which scored within the 22nd percentile as a first submission. Additionally, these data are being included in a manuscript that has been the product of an inter-SPORE collaboration between Stalnecker and Dr. Andrew Aguirre at the Dana-Farber Cancer Institute.
Calvin Cole, PhD (University of Rochester Medical Center): “Understanding Immune System Dysregulation and Potential for Therapy in PDAC- Induced Skeletal Muscle Wasting”
The goal of Cole’s project was to understand the role of canonical and non-canonical signaling of tumor-secreted IGFBP-3 in PDAC-induced skeletal muscle wasting. The findings from Cole’s experimentation in mice suggest that deletion of tumor-specific igfbp3 attenuates PDAC-induced skeletal muscle wasting (SMW) via downregulation of FoxO1 via both canonical and non-canonical signaling.
Canonically, tumor-associated IGFBP-3 blocks IGF-1 binding to its receptor, thus significantly inhibiting pAkt compared with a non-tumor control, and the activation of its downstream target mTOR. However, the deletion of tumor-associated igfbp3 significantly increases mTOR activation which is represented by increases in the ribosomal protein S6 (a downstream target of mTOR) when compared to mice that received parental PDAC cells.
Furthermore, findings from Cole’s lab demonstrate the possible signaling of tumor-associated IGFBP-3 via TGFβR1, which is significantly increased in PDAC mice compared to IGFBP-3(-/-). This increase in TGFβR1 signaling leads to the uninhibited activation of FoxO1, resulting in increases of genes associated with autophagy (maplc3b and ulk1) and the ubiquitin proteasome pathway (fbxo32 and trim63) in the muscle of PDAC mice compared to IGFBP-3(-/-) mice. Protein levels of autophagy-related LcrbII and Ulk1 are reduced in IGFBP-3(-/-) mice compared to PDAC mice.
Preliminary data showed that KCKO cells genetically deficient in IGFBP-3 (IGFBP-3-/-) fail to cause SMW in mice, in part due to down regulation of muscle-specific, ubiquitin proteasome pathway-associated E3 ligases and autophagy-related genes. Thus, Cole’s group sought to validate these findings in a murine pancreatic cancer model that recapitulates clinical pathology, specifically the CRISPR/Cas9 knockout of the igfbp3 gene in the murine KP2 PDAC cell line. The outcomes of this subsequent experiment yielded similar results to those of the lab’s initial findings. The Cole lab has most recently developed a novel neutralizing mAb against IGFBP-3 which is currently being tested in vivo.
—Tyler Rice, UNC Lineberger Pancreatic Cancer Center of Excellence