Argonaute 2 modulates EGFR-RAS signaling to promote mutant HRAS and NRAS-driven malignancies

Activating mutations in RAS GTPases drive nearly 30% of all human cancers. Our prior work described an essential role for Argonaute 2 (AGO2), of the RNA-induced silencing complex, in mutant KRAS-driven cancers. Here, we identified a novel endogenous interaction between AGO2 and RAS in both wild-type (WT) and mutant HRAS/NRAS cells. This interaction was regulated through EGFR-mediated phosphorylation of Y393-AGO2, and utilizing molecular dynamic simulation, we identified a conformational change in pY393-AGO2 protein structure leading to disruption of the RAS binding site. Knockdown of AGO2 led to a profound decrease in proliferation of mutant HRAS/NRAS-driven cell lines but not WT RAS cells. These cells demonstrated oncogene-induced senescence (OIS) as evidenced by β-galactosidase staining and induction of multiple downstream senescence effectors. Mechanistically, we discovered that the senescent phenotype was mediated via induction of reactive oxygen species. Intriguingly, we further identified that loss of AGO2 promoted a novel feed forward pathway leading to inhibition of the PTP1B phosphatase and activation of EGFR-MAPK signaling, consequently resulting in OIS. Taken together, our study demonstrates that the EGFR-AGO2-RAS signaling axis is essential for maintaining mutant HRAS and NRAS-driven malignancies.

Biochemical characterization of the interaction between KRAS and Argonaute 2

Oncogenic mutations in KRAS result in a constitutively active, GTP-bound form that in turn activates many proliferative pathways. However, because of its compact and simple architecture, directly targeting KRAS with small molecule drugs has been challenging. Another approach is to identify targetable proteins that interact with KRAS. Argonaute 2 (AGO2) was recently identified as a protein that facilitates RAS-driven oncogenesis. Whereas previous studies described the in vivo effect of AGO2 on cancer progression in cells harboring mutated KRAS, here we sought to examine their direct interaction using purified proteins. We show that full length AGO2 co-immunoprecipitates with KRAS using purified components, however, a complex between FL AGO2 and KRAS could not be isolated. We also generated a smaller N-terminal fragment of AGO2 (NtAGO2) which is believed to represent the primary binding site of KRAS. A complex with NtAGO2 could be detected via ion-mobility mass spectrometry and size exclusion chromatography. However, the data suggest that the interaction of KRAS with purified AGO2 (NtAGO2 or FL AGO2) is weak and likely requires additional cellular components or proteo-forms of AGO2 that are not readily available in our purified assay systems. Future studies are needed to determine what conformation or modifications of AGO2 are necessary to enrich KRAS association and regulate its activities.

Abstract 2578: An essential role for Argonaute 2 in HRAS and NRAS driven oncogenesis

The RAS gene family is among the most commonly mutated genes in cancer, and targeting the potential binding partners of mutated RAS may present a promising alternative therapeutic strategy. We recently identified Argonaute 2 (AGO2) of the RNA-induced silencing complex (RISC) as a novel partner of KRAS through its Switch II domain. Further, we demonstrated a role for AGO2 through the development of a KRASG12D driven mouse model of pancreatic cancer. Specifically, our findings revealed that AGO2 ablation drove early precursor pancreatic intraepithelial (PanIN) lesions to enter oncogene-induced senescence in lieu of progressing to late stage PanINs and pancreatic ductal adenocarcinoma. These studies suggested that AGO2 plays a key role in the development of mutant KRAS driven cancers.

In order to assess the role of AGO2 in other mutant RAS driven cancers, we performed co-IP of purified AGO2 and purified NRAS/HRAS proteins using isoform specific antibodies. Both HRAS and NRAS bound AGO2, regardless of GDP or GTP nucleotide loading. Using a panel of cell lines (including T24 cells (HRASG12V) and Mel-Juso cells (NRASQ61L)), we observed that AGO2 interacted with both wild-type (WT) and mutant forms of HRAS and NRAS. Additionally, using a variety of cell line models, we observed that EGFR-mediated phosphorylation of AGO2Y393 disrupted the interaction between WT RAS and AGO2. However, the mutant NRAS-, HRAS-, and KRAS-AGO2 interactions were resistant to EGFR regulation.

Considering the strong role of AGO2 in mutant KRAS driven cancer, we next asked if loss of AGO2 could induce a similar inhibition of RAS driven disease in human cell line models. In a panel of mutant NRAS and HRAS driven cell lines, AGO2 knockdown led to a substantial decrease in cell proliferation; however, WT RAS expressing cells were unperturbed following loss of AGO2. Similar to the loss of AGO2 in the KRASG12D driven mouse model, knockdown of AGO2 led to increased beta-galactosidase staining in a panel of oncogenic NRAS/HRAS driven cells. Furthermore, knockdown of AGO2 was associated with an induction of multiple senescence pathways, including increased p53, p21, and p16 expression. These results suggest that decreased AGO2 expression is sufficient to induce senescence in mutant but not WT RAS driven cells. In these cell line models, AGO2 loss reduced not only the mutant RAS isoform expression but also the WT RAS isoforms. Finally, knockdown of AGO2 led to an inhibition of T24 and Mel-Juso cell migration and metastasis in a zebrafish xenograft model. Taken together, these results suggest that 1) AGO2 interactions with mutant HRAS and NRAS play a key role in mutant RAS driven oncogenesis and 2) both mutant HRAS and NRAS depend on AGO2 to overcome senescence.

Citation Format: Ronald F. Siebenaler, Jessica Waninger, Sunita Shankar, Seema Chugh, Shih-Chun Chu, Jean Tien, Vijaya L. Dommeti, Malay Mody, Anudeeta Gautam, Carson Kenum, Chandan Kumar-Sinha, Arul M. Chinnaiyan. An essential role for Argonaute 2 in HRAS and NRAS driven oncogenesis [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 2578.

An essential role for Argonaute 2 in EGFR-KRAS signaling in pancreatic cancer development

Both KRAS and EGFR are essential mediators of pancreatic cancer development and interact with Argonaute 2 (AGO2) to perturb its function. Here, in a mouse model of mutant KRAS-driven pancreatic cancer, loss of AGO2 allows precursor lesion (PanIN) formation yet prevents progression to pancreatic ductal adenocarcinoma (PDAC). Precursor lesions with AGO2 ablation undergo oncogene-induced senescence with altered microRNA expression and EGFR/RAS signaling, bypassed by loss of p53. In mouse and human pancreatic tissues, PDAC progression is associated with increased plasma membrane localization of RAS/AGO2. Furthermore, phosphorylation of AGO2Y393 disrupts both the wild-type and oncogenic KRAS-AGO2 interaction, albeit under different conditions. ARS-1620 (G12C-specific inhibitor) disrupts the KRASG12C-AGO2 interaction, suggesting that the interaction is targetable. Altogether, our study supports a biphasic model of pancreatic cancer development: an AGO2-independent early phase of PanIN formation reliant on EGFR-RAS signaling, and an AGO2-dependent phase wherein the mutant KRAS-AGO2 interaction is critical for PDAC progression.

Abstract A21: Loss of Argonaute 2 leads to oncogene-induced senescence in mutant RAS-driven cancer

The RAS gene family is among the most commonly mutated genes within cancer, but little progress has been made in successfully targeting RAS mutations. Targeting binding partners of mutated RAS, however, presents a promising alternative therapeutic strategy. With the goal of uncovering novel interactors of RAS, we recently identified Argonaute 2 (AGO2) of the RNA-induced silencing complex (RISC) as a novel partner of the Switch II domain of KRAS. In order to assess the role of AGO2 in KRAS-G12D driven disease, we developed a mouse model of pancreatic cancer with conditional loss of AGO2. While AGO2 knockout did not prevent development of early precursor pancreatic intraepithelial (PanIN) lesions, loss of AGO2 prevented progression to late-stage PanINs, pancreactic ductal adenocarcinoma (PDAC), and metastatic disease. AGO2 null lesions displayed increased activation of the EGFR-RAS signaling axis during PanIN development. This signaling resulted in an increase in WT RAS-GTP activation, pEGFR-Y1068, and pERK levels leading to the development of oncogene-induced senescence in these PanIN lesions. Furthermore, we observed that EGFR-mediated phosphorylation of AGO2-Y393 disrupted the interaction between WT RAS and AGO2. This regulation by EGFR, however, was blocked in cells expressing mutant KRAS. These results suggested that the interaction of mutant RAS and AGO2 was vital to tumor development. To better assess the role of AGO2 loss in mutant RAS driven cancer, we performed AGO2 knockdown in multiple cell lines expressing mutations in either NRAS or HRAS isoforms. In each cell line, AGO2 directly interacted with KRAS, NRAS, and KRAS. In addition to suppressing growth in mutant RAS-driven cells (T24: HRAS-G12V, SK-MEL-2: NRAS-Q61H), loss of AGO2 produced marked increases in beta-galactosidase and p16 expression, as well as a decrease in cyclin D1, suggesting development of oncogene-induced senescence. Interestingly, upon AGO2 loss, cells displayed induction of pEGFR and pERK similar to what was observed in our pancreatic mouse model, and despite decreased expression of mutant RAS, WT RAS-GTP loading upon AGO2 loss was strongly induced. Together these results suggest a unique EGFR-AGO2-RAS signaling axis that requires AGO2-RAS interaction to prevent induction of oncogene-induced senescence in mutant RAS-driven cancers.

Citation Format: Ronald F. Siebenaler, Sunita Shankar, Jean C. Tien, Vijaya L. Dommeti, Sylvia Zelenka-Wang, Jessica Waninger, Malay Mody, Seema Chugh, Chandan Kumar-Sinha, Arul M. Chinnaiyan. Loss of Argonaute 2 leads to oncogene-induced senescence in mutant RAS-driven cancer [abstract]. In: Proceedings of the AACR Special Conference on Targeting RAS-Driven Cancers; 2018 Dec 9-12; San Diego, CA. Philadelphia (PA): AACR; Mol Cancer Res 2020;18(5_Suppl):Abstract nr A21.

Abstract A20: An essential role for Argonaute 2 in mouse models of KRAS driven cancers

In 2016, we identified a direct interaction between RAS and Argonaute 2 (AGO2), a key mediator of RNA-mediated gene silencing that is required for KRAS-driven oncogenesis using pancreatic and lung cancer cell line models. Recently, we employed the genetically engineered mouse model of pancreatic cancer to define the effects of conditional loss of AGO2 in KRASG12D driven pancreatic cancer. Genetic ablation of AGO2 did not interfere with development of the normal pancreas or KRASG12D-driven early precursor pancreatic intraepithelial neoplasia (PanIN) lesions. However, AGO2 loss prevents progression from early to late PanIN lesions, development of pancreatic ductal adenocarcinoma (PDAC), and metastatic progression. This results in a dramatic increase in survival of KRASG12D mutant mice deficient in AGO2 expression. Using validated pan-RAS and AGO2 antibodies for immunofluorescence (IF) and proximity ligation assay (PLA), we observed increased RAS and AGO2 co-localization at the plasma membrane in mouse and human pancreatic tissues associated with PDAC progression. AGO2 ablation permits PanIN initiation driven by the EGFR-RAS axis; however rather than progressing to PDAC, these lesions undergo profound oncogene-induced senescence (OIS). Since PanIN development requires EGFR and is not AGO2 dependent, we probed the effects of EGF stimulation in cell lines expressing wild-type and mutant forms of KRAS (using co-IP and PLA analyses). In wild-type RAS expressing cells, grown in media containing serum, RAS-AGO2 co-localization was limited to the intracellular regions of the cells, which dramatically increased and shifted to the plasma membrane under conditions of stress (serum starvation). Interestingly, EGF stimulation disrupted this membrane RAS-AGO2 interaction and restored it to intracellular basal levels. Using phosphorylation-deficient AGO2 mutants, we demonstrate that the disruption of wild type-RAS-AGO2 interaction is due to AGO2Y393 phosphorylation, a target of EGFR. Interestingly, the mutant KRAS-AGO2 interaction is not subject to EGFR activation, suggesting that although both the wild-type and mutant RAS bind AGO2, they are differentially regulated through growth factor receptor activation. Taken together, our study supports a biphasic model of pancreatic cancer development: an AGO2-independent early phase of PanIN formation reliant on EGFR and wild-type RAS signaling, and an AGO2-dependent phase wherein the mutant KRAS-AGO2 interaction is critical for PDAC progression. In the lung cancer mouse model, we also observed a similar dependence of AGO2 in KRAS-driven lung adenocarcinoma. Along with related abstracts detailing the mechanisms of OIS mediated by AGO2 (Ronald Siebenaler) and evidence of direct interaction between oncogenic KRAS and AGO2 with an affinity of 200nM (Jessica Waninger), we present our latest studies related to the KRAS-AGO2 interaction.

Citation Format: Sunita Shankar, Jean Ching-Yi Tien, Ronald F. Siebenaler, Vijaya L. Dommeti, Sylvia Zelenka-Wang, Jessica Waninger, Xiao-Ming Wang, Kristin M. Juckette, Alice Xu, Seema Chugh, Malay Mody, Sanjana Eyunni, Andrew Goodrum, Grace Tsaloff, Yuping Zhang, Ingrid J. Apel, Lisha Wang, Javed Siddiqui, Richard D. Smith, Heather A. Carlson, John Tesmer, Xuhong Cao, Jiaqi Shi, Chandan Kumar-Sinha, Arul M. Chinnaiyan. An essential role for Argonaute 2 in mouse models of KRAS driven cancers [abstract]. In: Proceedings of the AACR Special Conference on Targeting RAS-Driven Cancers; 2018 Dec 9-12; San Diego, CA. Philadelphia (PA): AACR; Mol Cancer Res 2020;18(5_Suppl):Abstract nr A20.

Abstract 957: An essential role for Argonaute 2 in mouse models of KRAS-driven cancers

In 2016, we identified a direct interaction between RAS and Argonaute 2 (AGO2), a key mediator of RNA-mediated gene silencing, that is essential for KRAS-driven oncogenesis using pancreatic and lung cancer cell line models. Recently, we employed a genetically engineered mouse model of pancreatic cancer to define the effects of conditional loss of AGO2 in KRASG12D-driven pancreatic cancer (KC model). Genetic ablation of AGO2 did not interfere with development of the normal pancreas or KRASG12D-driven early precursor pancreatic intraepithelial neoplasia (PanIN) lesions. However, AGO2 loss prevents progression from early to late PanIN lesions, development of pancreatic ductal adenocarcinoma (PDAC), and metastatic progression. This results in a dramatic increase in the survival of KRASG12D mutant mice deficient in AGO2 expression. Mechanistically, lack of PanIN to PDAC progression was due to oncogene-induced senescence (OIS) through activation of EGFR-wild type RAS-phosphoERK signaling in the absence of AGO2.

Using validated pan-RAS and AGO2 antibodies for immunofluorescence (IF) and proximity ligation assay (PLA), we observed increased RAS and AGO2 co-localization at the plasma membrane in mouse and human pancreatic tissues associated with PDAC progression. While AGO2 ablation permits PanIN initiation driven by the EGFR-RAS axis, these lesions undergo OIS rather than progressing to PDAC. Further, we used co-IP and PLA analyses to probe the effects of EGF stimulation in cell lines expressing wild-type and mutant forms of KRAS. In wild-type RAS expressing cells, RAS-AGO2 co-localization and interaction were limited to the intracellular regions of the cells, and dramatically increased and shifted to the plasma membrane under conditions of stress (serum starvation). Interestingly, EGF stimulation disrupted this membrane RAS-AGO2 interaction and restored it to intracellular levels. Using phosphorylation-deficient AGO2 mutants, we further demonstrate that the disruption of wild-type RAS-AGO2 interaction is due to EGFR-mediated AGO2Y393 phosphorylation. Interestingly, mutant KRAS-AGO2 interaction is not subject to EGFR activation, suggesting that although both the wild type and mutant RAS bind AGO2, they are differentially regulated through growth factor receptor activation.

We will discuss our ongoing studies evaluating the effects of AGO2 ablation in the KRASG12Ddriven lung cancer mouse model and PDAC progression with p53 loss (KPC model). Our recent in vivo work supports a biphasic model of pancreatic cancer development: an AGO2-independent early phase of PanIN formation reliant on EGFR and wild-type RAS signaling, and an AGO2-dependent phase wherein the mutant KRAS-AGO2 interaction is critical for PDAC progression.

Citation Format: Sunita Shankar, Jean Ching-Yi Tien, Ronald F. Siebenaler, Vijaya L. Dommeti, Sylvia Zelenka-Wang, Jessica Waninger, Kristin M. Juckette, Alice Xu, Xiao-Ming Wang, Seema Chugh, Malay Mody, Sanjana Eyunni, Andrew Goodrum, Grace Tsaloff, Yuping Zhang, Ingrid J. Apel, Javed Siddiqui, Richard D. Smith, Heather A. Carlson, John Tesmer, Xuhong Cao, Jiaqi Shi, Chandan Kumar-Sinha, Arul M. Chinnaiyan. An essential role for Argonaute 2 in mouse models of KRAS-driven cancers [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 957.

Abstract 94: Role of Argonaute 2 in oncogene induced senescence in a pancreatic cancer mouse model

Pancreatic cancer is a highly devastating malignancy with a very poor survival rate of 7%. Mutations in KRAS have been identified in more than 90% of PDAC patients. Previous work from our lab has shown that KRAS directly interacts with Argonaute 2 (AGO2) to promote cellular transformation. To study the involvement of AGO2 in KRASG12D-driven cancers, AGO2 expression was ablated in the KrasG12D/+; p48 Cre model (KC) model of pancreatic cancer. AGO2 floxed mice (AGO2loxP/loxP) were crossed with KC mice, resulting in oncogenic KRAS expression along with knockout of AGO2 in pancreatic acinar cells. Survival and disease progression were compared between wild-type (AGO2+/+; KRASG12D; p48Cre), heterozygous (AGO2fl/+; KRASG12D; p48Cre), and homozygous (AGO2fl/fl; KRASG12D; p48Cre) experimental mice groups. Homozygous knockout of AGO2 in KC mice resulted in significantly increased survival as compared to wild type and heterozygous mice. Pancreatic ductal adenocarcinoma (PDAC) and metastases were restricted to the wild-type and heterozygous mice. Pancreas from AGO2fl/fl; KRASG12D; p48Cre mice develop only early pancreatic intraepithelial lesions (PanINs), which fail to progress to PDAC. Senescence-associated β-galactosidase staining showed strong and significant increase in senescence in PanIN lesions mice lacking AGO2 expression as compared to AGO2+/+; KRASG12D; p48Cre mice. This suggests that AGO2 prevents oncogene induced senescence (OIS) as a result of KRASG12D expression and allows PanIN to PDAC progression.

To gain mechanistic insights of OIS due to AGO2 loss, we evaluated markers for OIS including p16, p53, p21, gamma γH2AX, and RAS-associated signaling (pERK and pAkt). Analysis of PanIN lesions lacking AGO2 showed increased p16 levels and high levels of phospho-ERK, compared to PDAC from pancreas with AGO2 expression. In order to extend these observations in cell line models, we performed AGO2 knockdown in T24 cells harboring HRASG12V. Surprisingly, cells with low AGO2 levels underwent OIS, which was similar to the pancreatic mouse model and was accompanied with increased phospho-ERK signaling and p16 expression. Further studies are underway to determine the contribution of the RAS-AGO2 interaction in the development of OIS.

Additionally, we are using CRISPR/Cas9 technology to screen pancreatic cancer cell lines with AGO2 knockout for their dependence on AGO2 and their ability to undergo OIS in the absence of AGO2 expression. We will present findings from our ongoing studies involving the role of AGO2 loss in the KPC (KRASG12D; p53fl/+; Cre) model, wherein OIS will be assessed in the absence of p53, a canonical inducer of cellular senescence.

Citation Format: Seema Chugh, Jean C. Tien, Ronald F. Siebenaler, Vijaya L. Dommeti, Sylvia Z. Wang, Sanjana Eyunni, Kristin M. Juckette, Lisha Wang, Sunita Shankar, Arul M. Chinnaiyan. Role of Argonaute 2 in oncogene induced senescence in a pancreatic cancer mouse model [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 94.

Abstract 956: An essential role for Argonaute 2 in EGFR-KRAS signaling in pancreatic cancer development

The RAS gene family is among the most commonly mutated genes within cancer, but little progress has been made in successfully targeting RAS mutations. Targeting binding partners of mutated RAS presents as a promising alternative therapeutic strategy. With the goal of uncovering novel interactors of RAS, we recently identified Argonaute 2 (AGO2) of the RNA-induced silencing complex (RISC) as a novel partner of the KRAS through its Switch II domain. In order to assess the role of AGO2 in KRASG12D driven disease, we developed a mouse model of pancreatic cancer with conditional loss of AGO2. While AGO2 knockout did not prevent development of early precursor pancreatic intraepithelial (PanIN) lesions, AGO2 null lesions displayed increased activation of the EGFR-RAS signaling axis during PanIN development that failed to progress to late stage PanINs, pancreatic ductal adenocarcinoma (PDAC), and metastatic disease. This resulted in a dramatic increase in the survival of mice with AGO2 ablation. Unlike the PanINs in AGO2 sufficient mice, the early PanIN lesions with AGO2 ablation showed staining for the senescence associated beta galactosidase activity, suggesting that AGO2 loss induces oncogene induced senescence. To extend these observations and explore the role of AGO2 interaction with mutant forms of HRAS and NRAS proteins, we performed co-IP of AGO2 with RAS proteins using isoform specific antibodies. Both HRAS and NRAS bound AGO2 in T24 cells (HRASG12V) and SK-MEL-2 cells (NRASQ61H), respectively. In T24 cells, AGO2 knockdown led to the senescent phenotype and was accompanied with changes in the EGFR-RAS signaling axis, similar to that observed in the PanINs of the mice with AGO2 loss. In this cell line model, AGO2 loss reduced mutant HRAS expression and increased wild type RAS activity. These signaling effects were also consistent with our observation that AGO2 loss increased RAS activation in the mouse embryonic fibroblast (MEF) model. Together with our previous work with mutant KRAS dependent cells, these data suggest that 1) AGO2-wild type RAS binding prevents RAS activation and 2) mutant RAS-AGO2 association regulates oncogenic RAS levels in cell line models. Studies on the mouse model and the close proximity of RAS and AGO2 with EGFR also furthered our understanding of the RAS-AGO2 interaction. Using a variety of cell line models, we observed that EGFR-mediated phosphorylation of AGO2Y393 disrupts the interaction between WT RAS and AGO2. However, the mutant KRAS-AGO2 interaction was recalcitrant to EGFR regulation. This provides the first instance of a nucleotide dependent association of RAS and AGO2 and sheds light on the dynamic nature of the RAS-AGO2 interaction.

Citation Format: Ronald F. Siebenaler, Sunita Shankar, Jean C. Tien, Vijaya L. Dommeti, Sylvia Zelenka-Wang, Seema Chugh, Ingrid J. Apel, Malay Mody, Anudeeta Gautam, Chandan Kumar-Sinha, Arul M. Chinnaiyan. An essential role for Argonaute 2 in EGFR-KRAS signaling in pancreatic cancer development [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 956.

Abstract 3020: An essential role for Argonaute 2 in EGFR-KRAS signaling in pancreatic cancer development

KRAS and EGFR have been shown to function as essential mediators of pancreatic cancer development. In addition, KRAS and EGFR have been separately shown to interact with and perturb the function of Argonaute 2 (AGO2), a key mediator of RNA-mediated gene silencing. Here, we employed a genetically engineered mouse model of pancreatic cancer to define the effects of conditional loss of AGO2 in KRASG12D driven pancreatic cancer. Genetic ablation of AGO2 does not interfere with development of the normal pancreas or KRASG12D driven early precursor pancreatic intraepithelial (PanIN) lesions. However, AGO2 loss prevents progression from early to late PanIN lesions, development of pancreatic ductal adenocarcinoma (PDAC), and metastatic progression. This results in a dramatic increase in survival of KRASG12D mutant mice deficient in AGO2 expression. In both mouse and human pancreatic tissues, increased AGO2 expression at the plasma membrane is associated with PDAC progression. Mechanistically, within early precursor PanIN lesions, loss of AGO2 elevates phospho-EGFR levels and activates wild-type RAS, antagonizing KRASG12D activation and PDAC development. Furthermore, we observe that phosphorylation of AGO2Y393 by EGFR disrupts the interaction of wild-type RAS with AGO2, but does not affect the interaction of mutant KRAS with AGO2. Taken together, our study supports a biphasic model of pancreatic cancer development: an AGO2-independent early phase of PanIN formation reliant on EGFR and wild-type RAS signaling, and an AGO2-dependent phase wherein the KRAS-AGO2 interaction is critical to the progression from PanIN to PDAC. A recent study by Phillip Sharp and Tyler Jacks describes a requirement for the oncogenic KRAS-AGO2 interaction in the development of a transplant mouse model of plasmablastic lymphoma, and together these studies substantiate the role of the interaction in KRAS oncogenesis.

Citation Format: Sunita Shankar, Jean Tien, Ronald F. Siebenaler, Vijaya L. Dommeti, Sylvia Zelenka-Wang, Kristin M. Juckette, Alice Xu, Malay Mody, Andrew Goodrum, Grace Tsaloff, Ingrid J. Apel, Lisha Wang, Javed Siddiqui, Jiaqi Shi, Chandan Kumar-Sinha, Arul Chinnaiyan. An essential role for Argonaute 2 in EGFR-KRAS signaling in pancreatic cancer development [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 3020.

Abstract 4370: Regulation of AGO2-KRAS interaction through epidermal growth factor receptor

The RAS gene family is among the most commonly mutated genes within cancer. While much research has elucidated its major functions and downstream pathways, little progress has been made in successfully targeting RAS mutations. We recently identified Argonaute 2 (AGO2) of the RNA-induced silencing complex (RISC) as a novel partner of the Switch II domain of KRAS. We have found that stable knockdown of AGO2 in KRAS-dependent cell lines lead to a decrease in KRAS protein expression with a subsequent decrease in cellular proliferation. In addition, we observed a decrease in microRNA unwinding in the presence of mutant KRAS, suggesting this interaction inhibits the endogenous RNAi function of AGO2. Despite this clear connection between KRAS and AGO2 in KRAS mediated oncogenesis, the precise function of this interaction remains unclear in both normal and cancer biology.

In order to identify endogenous regulators of AGO2-RAS, we investigated the ability of EGFR signaling to modulate the AGO2-RAS interaction. We established two mouse embryonic fibroblast cell lines (NIH-3T3 and MEF) with complete knockout of AGO2. When compared to normal control cells, we found that knockout of AGO2 resulted in an increase in WT RAS-GTP activation levels, phosphorylation of Y1068-EGFR, and MAPK/ERK and PI3K/AKT signaling. Rescue of AGO2 knockout resulted in a return to normal levels of active RAS-GTP, pEGFR, and downstream signaling.

Recent studies have described EGFR phosphorylation of AGO2 under hypoxic cell conditions, resulting in the inhibition of AGO2 association with RISC members. In order to better characterize the relationship between EGFR-AGO2-RAS, we found that overnight serum starvation followed by stimulation with EGF led to a decrease in AGO2-RAS co-IP in WT KRAS cells. Blocking AGO2 phosphorylation with a Y393F mutant of AGO2 prevented AGO2-RAS dissociation following EGFR stimulation. While WT KRAS cell lines displayed regulation of AGO2-RAS via EGFR, the phosphorylation of AGO2Y393 was unable to disrupt the interaction of AGO2 with mutant KRAS following stimulation with EGF. Together these results suggest a unique EGFR-AGO2-RAS signaling axis, and its dysregulation by mutant KRAS could increase oncogenic growth through promotion of AGO2-RAS interaction in cancer.

Citation Format: Ronald F. Siebenaler, Sunita Shankar, Jean C. Tien, Vijaya L. Dommeti, Malay Mody, Chandan Kumar-Sinha, Arul M. Chinnaiyan. Regulation of AGO2-KRAS interaction through epidermal growth factor receptor [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 4370.

Abstract 1362: Argonaute 2 controls RAS activation in mouse embryonic fibroblasts

The RAS gene family is among the most commonly mutated genes within cancer, and while much research has elucidated the major downstream pathways, including MAPK and PI3K, little progress has been made in successfully targeting mutant RAS in cancer. We recently identified an interaction between the N terminal domain of Argonaute 2 (AGO2), a core component of RNA-induced silencing complex (RISC), and the Switch II domain of KRAS. Furthermore, this interaction was found in all cell lines tested, expressing either wild-type (WT) or mutant KRAS. We found that stable knockdown of AGO2 in KRAS dependent cell lines lead to a decrease in KRAS protein expression with a subsequent decrease in cellular proliferation. Conversely, the overexpression of AGO2 in these cells lead to both an increase in KRAS expression and oncogenesis. In addition, this interaction inhibits the RNAi function of AGO2 by preventing microRNA unwinding in the presence of oncogenic KRAS compared to WT-KRAS. Despite a clear association between mutant KRAS and AGO2 mediating increased KRAS mediated oncogenesis, the precise function of this interaction remains unclear in normal physiology. In order to better assess the endogenous function of the KRAS-AGO2 interaction, we analyzed two mouse embryonic fibroblast cell lines (NIH 3T3 and MEF) with complete knockout of AGO2. Utilizing a Raf-1 RAS binding domain (RBD) pulldown method, we assessed activated WT-RAS levels in AGO2 null NIH 3T3 and MEF cells. We found that knockout of AGO2 lead to an increase in WT RAS-GTP activation compared to normal control cells. Immunoblot analysis also indicates that AGO2 null fibroblasts lead to increase in RAS downstream signaling through the MAPK/ERK and PI3K/AKT pathways. Furthermore, rescue of AGO2 knockout using full length mouse AGO2 decreased wild type RAS activation and its downstream signaling. Taken together, these observations suggest that the AGO2 interaction may suppress WT-KRAS activation, leading to maintenance of RAS-GDP levels. Using RNA-seq, proteome and microRNA analysis, we have begun to identify the pathways that may be involved in RAS activation in AGO2 null cells. Early analyses indicate that AGO2 controls WT-KRAS levels and activity through multiple mechanisms, laying the foundation for a better understanding of the RAS-AGO2 interaction in normal physiology.

Citation Format: Ronald F. Siebenaler, Sunita Shankar, Vijaya L. Dommeti, Malay Mody, Arul Chinnaiyan. Argonaute 2 controls RAS activation in mouse embryonic fibroblasts [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 1362. doi:10.1158/1538-7445.AM2017-1362

Abstract 3849: The partial tandem duplication of Mll (Mll PTD) is a gain-of-function in the absence of Mll wildtype (Mll WT) in adult mouse hematopoiesis

Cytogenetically normal acute myeloid leukemia (CN AML) patients with a MLL PTD have a poor prognosis compared to CN AML patients with MLL WT. We previously reported that the MLL WT gene in MLL PTD+ CN AML is epigenetically silenced. To investigate in vivo significance of the MLL PTD in the absence of MLL WT on hematopoiesis and leukemogenesis, we generated homozygous MllPTD/PTD and hemizygous MllPTD/- mice. These mice died in utero or as neonates, respectively, precluding further study on adult hematopioesis and leukemogenesis. In the current study, we crossed MllPTD/WT mice with Mll-conditional knock-out (cKO) animals to produce MllPTD/cKO mice. Like the MllPTD/WT mice, the MllPTD/cKO mice survive to adulthood, as both models express Mll WT and PTD. As previously reported by Jude et al, (Cell Stem Cell, 2007) Cre activation in hematopoietic cells of mice carrying the Mll cKO allele resulted in an intragenic deletion in Mll (deltaN allele) that when expressed, the protein was unable to translocate to the nucleus and thereby unable to exert normal Mll function. DNA PCR confirmed the correct MllPTD/deltaN genotype was generated. MlldeltaN/deltaN mice develop bone marrow (BM) failure (marked hypocellularity) at a median of 18 days post-Cre activation. Comparatively, MllPTD/deltaN mice (n=5) survived beyond 52 weeks post-Cre activation (P<0.001), without evidence of BM failure. Also, at one year, MllPTD/deltaN animals exhibit equivalent white blood cell counts and hematocrit when compared to MllWT/WT, MllPTD/WT and MllPTD/cKO controls. To evaluate whether Mll PTD function was maintained in bone marrow, qRT-PCR analysis of the Mll transcriptional target Hoxa9 at 12 days post-Cre activation revealed a 20-fold reduction in Hoxa9 mRNA levels in MlldeltaN/deltaN animals and ∼4-5 fold increase in MllPTD/deltaN and MllPTD/cKO BM (P=0.025) compared to WT BM. MllPTD/cKO animals, with two functional copies of Mll, had nearly equivalent overexpression of HoxA9 as seen in MllPTD/deltaN mice that maintain only one functional copy of Mll, supporting our previous hypothesis that the Mll PTD acts as a gain-of-function mutation. In conclusion, these data demonstrate for the first time that Mll PTD alone can support adult hematopoiesis. In conclusion, these data demonstrate for the first time that Mll PTD alone can support adult hematopoiesis and even though Hoxa9 is upregulated, another oncogene is required for leukemogenesis. Absence of Mll WT is not sufficient to promote acute leukemogenesis in Mll PTD+ mice, but appears to be a contributory factor. We recently reported that MllPTD/WT mice crossed with Flt3ITD/WT mice, the latter of which also do not develop AML (Lee et al. Cancer Cell, 2006), generated double mutant mice that do develop AML. Leukemic blasts from these mice exhibit a reduction in Mll WT expression (Zorko et al, Blood, 2012), recapitulating what is seen in human MLL PTD+ CN AML.

Citation Format: Nicholas A. Zorko, Daniel A. Yanes, W. Courtland Lewis, Daniel L. Brook, Kelsie M. Bernot, Ronald F. Siebenaler, Susan P. Whitman, Elshafa H. Ahmed, Kathleen K. McConnell, John Nemer, Patricia Ernst, Gang Huang, Guido Marcucci, Michael A. Caligiuri. The partial tandem duplication of Mll (Mll PTD) is a gain-of-function in the absence of Mll wildtype (Mll WT) in adult mouse hematopoiesis. [abstract]. In: Proceedings of the 104th Annual Meeting of the American Association for Cancer Research; 2013 Apr 6-10; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2013;73(8 Suppl):Abstract nr 3849. doi:10.1158/1538-7445.AM2013-3849

Abstract 3249: Targeting the miR29b/Sp1/Dnmt pathway for curative therapy in Mll-PTD/Flt3-ITD acute myeloid leukemia

The partial tandem duplication (PTD) of the MLL gene is a molecular defect in human myelodysplastic syndrome and acute myeloid leukemia (AML) and associates with poor patient outcomes, especially when other molecular prognostic markers are co-present such as the internal tandem duplication (ITD) of FLT3, i.e., FLT3-ITD. In human AML, MLL-PTD alters the epigenome via aberrant H3K4 methylation activity but is also associated with increased DNA methylation through an unknown mechanism (Whitman, et al, Blood 2008). Using a double knock-in mouse model of AML with Mll-PTD and Flt3-ITD, we analyzed the DNA methylome using a methyl binding protein (MBD2) pull down followed by next generation sequencing. Leukemic mouse bone marrow (BM, n=8) exhibited a 1.7 fold increase in global DNA methylation compared to non-leukemic controls (n=9, P=0.017). Consistent with this finding, DNA methyltransferases (Dmnt) 1, 3a, and 3b were overexpressed (1.5 fold P=0.03, 2.3 fold P=0.01, and 5.3 fold P=0.03, respectively). We previously showed that the non-coding microRNA 29b (miR-29b) expression directly downmodulates DNMT3a and 3b, but indirectly inhibits DNMT1 via miR-29b-driven SP1 suppression. In the murine Mll-PTD/Flt3-ITD AML, the precursor to the mature miR-29b (pri-miR-29b-2) was downregulated in leukemic BM (0.7-fold, P<0.001), whereas Sp1 mRNA was upregulated compared to wildtype (1.6-fold P=0.01). Bortezomib is a proteasome inhibitor that also downregulates Dnmt's (Liu, et al, Blood, 2008) by increasing the expression of miR-29b (Liu, et al, Cancer Cell, 2010). In vitro, bortezomib induced a dose-dependent reduction in proliferation of Mll-PTD/Flt3-ITD leukemic blasts (MTS assay, EC50 23.83nM) and apoptosis. Prior to cell death in bortezomib-treated cells, expression of pri-miR-29b-2 increased by 1.7 fold over vehicle treated cells (P=0.01) while Dnmt3b mRNA (P=0.001) and protein levels decreased to half of the vehicle treated level. Targeting of this pathway in vivo using liposomal-bortezomib (1mg/kg IV twice per week for 2 weeks followed by 2mg/kg IV twice per week for 2 weeks) significantly increased survival of AML transplanted mice, compared with vehicle or free-bortezomib with 80% alive 90 days post-transplant (P<0.001). At day 90, treated mice had normal spleen weights (mean 86mg) and absence of leukemic blasts in BM, liver, or spleen compared to vehicle-treated mice that showed significant blast infiltration (mean spleen weight 396mg, P<0.001). Taken together, these data support the miR-29b/SP1/DNMT pathway of epigenetic modulation as active in Mll-PTD/Flt3-ITD AML, and indicate bortezomib as a valuable treatment option to be explored in patients with MLL-PTD/FLT3-ITD-associated AML.

Citation Format: Kelsie M. Bernot, John S. Nemer, Ramasamy Santhanam, Shujun Liu, Gabriel G. Marcucci, Nicholas A. Zorko, Susan P. Whitman, Pearlly Yan, David Frankhouser, Ralf Bundschuh, Mengzi Zhang, Ronald F. Siebenaler, Elshafa H. Ahmed, Kathleen K. McConnell, Maura Munoz, Daniel L. Brook, Adrienne M. Dorrance, Katy E. Dickerson, Xiaoli Zhang, Jianying Zhang, David Jarjoura, Robert Lee, William Blum, Michael A. Caligiuri, Guido Marcucci. Targeting the miR29b/Sp1/Dnmt pathway for curative therapy in Mll-PTD/Flt3-ITD acute myeloid leukemia. [abstract]. In: Proceedings of the 104th Annual Meeting of the American Association for Cancer Research; 2013 Apr 6-10; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2013;73(8 Suppl):Abstract nr 3249. doi:10.1158/1538-7445.AM2013-3249