Targeting integrin-linked kinase to suppress oncogenic KRAS signaling in pancreatic cancer

ABSTRACT

Although oncogenic KRAS represents a therapeutically relevant target in pancreatic cancer, it is deemed “non-druggable” because of the intrinsic difficulty in designing direct inhibitors of KRAS. Our recent work demonstrated a KRAS-integrin-linked kinase (ILK) regulatory feedback loop that allows pancreatic cancer cells to regulate KRAS expression and to interact with the tumor microenvironment to promote aggressive phenotype. KRAS induces E2F1-mediated transcriptional activation of ILK expression, and ILK, in turn, controls KRAS expression via hnRNPA1, which binds and destabilizes the G-quadruplex in the KRAS promoter. Moreover, ILK inhibition blocked KRAS-driven EMT and growth factor-stimulated KRAS expression. This regulatory loop, however, was not noted in KRAS mutant colorectal and lung cancer cells examined as knockdown of KRAS or ILK did not affect each other's expression, suggesting that this KRAS-ILK feedback regulation is specific for pancreatic cancer. In sum, this regulatory loop provides a strong mechanistic rationale for suppressing oncogenic KRAS signaling through targeting ILK, and this creating a potential new therapeutic strategy for pancreatic cancer.

Pancreas cancer remains a formidable challenge with very little, if any, advance in survival over the last few decades.¹ A consensus report from the NCI Clinical Trials Planning Meeting on Pancreas Cancer Treatment placed a strong emphasis on the importance of KRAS targeting as a top priority that should continue to be pursued.² Activating mutations (point mutations at codons G12, G13, and Q61) in the KRAS oncogene are the most frequent genetic abnormality in more than 90% of human pancreatic cancers.³ KRAS cycles between an inactive GDP-bound “off” state and an active GTP-bound “on” state, and functions as a molecular switch that mediates signal transduction between membrane growth factor receptors and downstream pathways governing oncogenesis, metastasis, and tumor progression, among which the PI3K/Akt and RAF-MEK-ERK pathways are most noteworthy (Fig. 1).⁴ Mechanistically, these oncogenic mutations result in the accumulation of KRAS in the active GTP-bound state, leading to constitutive activation of downstream signaling pathways. Transgenic mouse models that expressed an oncogenic KRAS allele (KRAS^G12D or KRAS^G12V) in pancreas have shown that such KRAS mutations not only play a crucial role in the initiation of pancreas carcinogenesis,^5-7 but also are required for tumor maintenance.^8,9 From a therapeutic perspective, oncogenic KRAS represents a clinically relevant target in pancreatic cancer. However, previous attempts to block the posttranslational modifications or membrane attachment of KRAS have not been successful clinically. Moreover, it proves challenging to design direct RAS inhibitors as RAS proteins lack intrinsic druggable target sites that allow small molecules to effectively compete.^10-12 Nevertheless, recent advances have been made in developing irreversible inhibitors of G12C mutant KRAS via structure-based design, which provides a proof-of-concept that oncogenic KRAS might be directly targeted through allosteric inhibition.^13-15 In addition, allosteric inhibitors of the RAS-like GTPase Ral have also been reported, which act by selective binding to a site on the GDP-bound form of Ral.¹⁶

Figure 1.

Schematic diagram depicting KRAS signaling in pancreatic cancer cells.

From a translational perspective, an alternative approach to target oncogenic KRAS is to suppress its expression via pharmacological inhibition of an upstream regulator. For example, the promoter of the KRAS protooncogene contains a critical nuclease hypersensitive element forming a parallel G-quadruplex (also known as G4-DNA) structure,¹⁷ which is a higher order DNA structure formed from guanine (G)‐rich.¹⁸ This G-quadruplex is recognized by several nucleoproteins, among which heterogeneous nuclear ribonucleoprotein A1 (hnRNPA1) is noteworthy in its ability to destabilize and relax the G-quadruplex structure to facilitate KRAS gene transcription.¹⁹ Mechanistically, identification of small-molecule agents capable of stabilizing this G-quadruplex structure in the upstream region essential for the promoter activity of KRAS gene will result in the downregulation of its gene expression. The proof-of-concept of this G4-targeting strategy was obtained by G4-mimicking oligonucleotides (G4-decoys), which could bind to and stabilize one of the G4 structures in the 5′UTR of KRAS mRNA, resulting in the suppression of KRAS protein expression and cell growth in pancreatic cancer cells.²⁰

Recently, we reported a novel function of integrin-linked kinase (ILK) in regulating the expression of KRAS through an autoregulatory loop in KRAS mutant pancreatic cancer cells.²¹ ILK is a serine/threonine kinase with diverse oncologic functions,^22,23 which has been associated with the regulation of pancreatic cancer proliferation, adhesion and invasion, and epithelial–mesenchymal transition (EMT).^24-26 We obtained evidence that oncogenic KRAS upregulates ILK expression through E2F1-facilitated transcriptional activation, and ILK, in turn, mediates KRAS signaling in 2 ways (Fig. 1).

First, ILK contributes to the maintenance of oncogenic KRAS expression. Specifically, ILK increases hnRNPA1 expression via c-Myc upregulation, which, in turn, facilitates KRAS transcription by destabilizing the G-quadruplex on the KRAS promoter. Mechanistically, this newly identified role of hnRNPA1 as a link between ILK and oncogenic KRAS is noteworthy as it not only regulates the expression of KRAS and other oncogenic proteins, but also has diverse functions in mRNA biogenesis and processing, telomere maintenance and the regulation of transcription factor activity.²⁷ Second, ILK facilitates tumor progression and metastasis, in part, by upregulating YB-1 and Twist expression.²⁸ Substantial evidence indicates that Twist and the YB-1 target, Snail, are master regulators of EMT.^29,30 Accordingly, genetic knockdown or pharmacological inhibition of ILK reversed the mesenchymal phenotypes of pancreatic cancer cells. Together, these findings suggest that ILK might, in part, be responsible for the effect of oncogenic KRAS on EMT and other aggressive phenotype.

Equally important, our study also suggests the potential involvement of this regulatory loop in regulating the crosstalk between growth factor receptor signaling (EGFR and insulin-like growth factor 1 receptor) and oncogenic KRAS (Fig. 1). Although EGFR signals mostly through KRAS by increasing its activity, inhibition of EGFR is expected to have little or no effect on oncogenic KRAS-driven signaling pathways due to their constitutively active status. However, recent evidence indicates that EGFR signaling is still essential for oncogenic KRAS-driven pancreatic tumorigenesis.^31,32 Mechanistically, the ability of EGF to upregulate oncogenic KRAS expression might underlie this EGFR-dependency. Moreover, it is intriguing that insulin is able to upregulate KRAS expression, which might explain the reported epidemiological link between higher insulin concentrations and increased pancreatic cancer risk.³³ The clinical implication of the functional role for this regulatory loop in facilitating the crosstalk between oncogenic KRAS and the tumor microenvironment in pancreatic cancer warrants further investigations.

Pursuant to the above findings, we raised a question of whether this KRAS-ILK regulatory loop was also functional in other types of cancer cells, and thus examined the effect of KRAS knockdown on ILK expression, and vice versa, in several KRAS mutant colorectal and lung cancer cell lines, including HCT-116, SW480, H157, and A549. In contrast to pancreatic cancer cells, silencing of KRAS or ILK in these cell lines had no appreciable effect on each other's expression (Fig. 2), refuting the involvement of ILK in regulating oncogenic KRAS expression in these cancer cells.

Figure 2.

Effect of siRNA-mediated knockdown of KRAS on the expression of ILK, and vice versa, in HCT-116 and SW480 colon cancer and H157 and A549 lung cancer cells.

We rationalize that the specificity of this KRAS-ILK loop in pancreatic cancer cells might be attributable to differences in the mechanisms that underlie the regulation of the expression of the 2 key intermediary effectors E2F1 and hnRNPA1 in different types of cancer cells. For example, it has been reported that the lysine acetyltransferase GCN5 plays a critical role in regulating E2F1 expression in lung cancer cells,³⁴ and that hnRNPA1 is negatively regulated by miR-18a in colon cancer cells though the induction of autophagolysomal degradation.³⁵ Consequently, it is plausible that lung and colon cancer cells might adopt alternative mechanisms to regulate the expression of E2F1 and hnRNPA1, thereby disrupting the interplay between KRAS and ILK in forming a regulatory feedback loop.

In addition, this discrepancy is reminiscent with the differential ability of ILK to facilitate the phosphorylation of Akt at Ser-473 in different cell lines. Although ILK has been reported to act as phosphoinositide-dependent kinase (PDK)-2 to facilitate the phosphorylation of Ser-473-Akt in many cancer cell lines,^22,23 our data showed that none of the KRAS mutant pancreatic cancer cell lines examined, including AsPC-1, Panc-1, and BxPC-3, were susceptible to the suppressive effect of ILK knockdown on Ser-473-Akt phosphorylation (not shown). Together, these findings underlie the cell type- and/or context dependency of ILK in regulating different oncogenic functions in cancer cells.

Considering the diverse function of ILK in regulating metastasis and aggressive phenotype of KRAS mutant pancreatic cancer cells, the pharmacological exploitation of the ability of ILK to regulate KRAS expression by developing potent ILK inhibitors is a viable therapeutic strategy of pancreatic cancer, which constitutes the focus of our current investigation.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.

Funding

This work was supported by the Lucius A Wing Endowed Chair Fund (CSC), Ministry of Science and Technology grant MOST 103-2320-B-001-017 (PCC), Bi-institutional Collaborative Pancreatic Cancer Research grants from National Cheng Kung University College of Medicine (CSC) and The Ohio State University Wexner Medical Center and Comprehensive Cancer Center (TBS).