Pancreatic Tumorigenesis: Oncogenic KRAS and the Vulnerability of the Pancreas to Obesity

Simple Summary Pancreatic cancer is a devastating disease with a poor survival rate, and oncogenic mutant KRAS is a major driver of its initiation and progression; however, effective strategies/drugs targeting major forms of mutant KRAS have not been forthcoming. Of note, obesity is known to worsen mutant KRAS-mediated pathologies, leading to PDAC with high penetrance; however, the mechanistic link between obesity and pancreatic cancer remains elusive. The recent discovery of FGF21 as an anti-obesity and anti-inflammation factor and as a downstream target of KRAS has shed new light on the problem. Abstract Pancreatic ductal adenocarcinoma (PDAC) is one of the most lethal malignancies and KRAS (Kirsten rat sarcoma 2 viral oncogene homolog) mutations have been considered a critical driver of PDAC initiation and progression. However, the effects of mutant KRAS alone do not recapitulate the full spectrum of pancreatic pathologies associated with PDAC development in adults. Historically, mutant KRAS was regarded as constitutively active; however, recent studies have shown that endogenous levels of mutant KRAS are not constitutively fully active and its activity is still subject to up-regulation by upstream stimuli. Obesity is a metabolic disease that induces a chronic, low-grade inflammation called meta-inflammation and has long been recognized clinically as a major modifiable risk factor for pancreatic cancer. It has been shown in different animal models that obesogenic high-fat diet (HFD) and pancreatic inflammation promote the rapid development of mutant KRAS-mediated PDAC with high penetrance. However, it is not clear why the pancreas with endogenous levels of mutant KRAS is vulnerable to chronic HFD and inflammatory challenges. Recently, the discovery of fibroblast growth factor 21 (FGF21) as a novel anti-obesity and anti-inflammatory factor and as a downstream target of mutant KRAS has shed new light on this problem. This review is intended to provide an update on our knowledge of the vulnerability of the pancreas to KRAS-mediated invasive PDAC in the context of challenges engendered by obesity and associated inflammation.


Introduction
PDAC is the major histologic subset of pancreatic cancer, accounting for more than 90% of all pancreatic malignancies. Despite advances in medical knowledge that has improved the survival rate of many types of aggressive cancer, PDAC remains one of the deadliest malignancies with an incidence/mortality ratio of as high as 94% [1,2] and a five-year survival rate of about 9% [3]. It is predicted that within a decade PDAC will rise to the second-leading cause of cancer-related deaths in the U.S. [4,5]. Even more threatening, there are currently no highly effective standard therapies for the large majority rats chronically fed a HFD developed signs of the fatty pancreas, pancreatic injury, and fibrosis with significantly elevated levels of FFAs, triglycerides, insulin, and monocyte chemoattractant protein-1. Consistent with these observations, clinical studies in PDAC patients showed a significant correlation between increased serum IGF1 concentration and poorer survival [58].
White adipose tissue also secretes other so-called adipokine soluble factors, notably leptin and adiponectin, among others. Leptin functions by binding and activating the leptin receptor (Ob-R), a member of the class I cytokine receptor family, and through cross-talk with the Notch receptor pathways and signals satiety to the hypothalamus to regulate dietary intake, energy expenditure, and fat storage [59]. Obesity is associated with leptin resistance despite increased leptin levels [60]. Case-control studies showed that high plasma leptin concentrations were associated with an increased risk of developing PDAC [61,62]. Obese conditions induce hypoxia and activate hypoxia-inducible factor 1alpha (HIF1a), which was shown to directly activate Ob-R in pancreatic cancer cells [63]. A growing body of experimental evidence suggests that increased leptin signaling promotes pancreatic cancer progression, metastasis, and chemoresistance [64,65]. On the other hand, low serum concentrations of adiponectin, another unique adipose-derived adipokine, was found inversely associated with increased visceral fat in overweight and obese subjects and was associated with worse outcomes for PDAC patients [66,67]. Kras G12D/+ expressing mice exposed to intermittent and chronic calorie-restricted diets showed a reduced incidence of pancreatic intraepithelial neoplasia (PanIN) and delayed PDAC progression accompanied by increased serum adiponectin and decreased serum leptin levels [57].
Pathological expansion of visceral adiposity and accumulation of fat in obese conditions are also known to evoke pro-inflammatory or inflammatory cellular changes characterized by a pattern of prominent infiltration of macrophages, T helper type 1 (Th1) cells, neutrophils, monocytes, and natural killer cells, as well as production and secretion of a plethora of pro-inflammatory cytokines, such as interleukin 6 (IL6), tumor necrosis factor alpha (TNFα), and interleukin 1 alpha (IL1α). These changes can occur in the peripheral adipose tissues, peri-pancreatic fat, and the pancreas or pancreatic tumor foci, contributing to PDAC development and invasiveness [22,[68][69][70]. In turn, increased IL6, TNFα, and IL1α levels were showed to significantly activate c-Jun N-terminal kinases, nuclear factor-κB (NFκB), signal transducer and activator of transcription 3, Janus kinase 2, and MAPK signaling pathways, promoting inflammation, stromal cell activation, and pancreatic cancer cell growth. Deletion of tumor necrosis factor receptor 1 or IL6 in Kras G12D/+ -expressing mice significantly reduced HFD-or inflammation-promoted tissue damage, fibrosis, cancer cell proliferation, and PanIN development [50,71]. Furthermore, these pro-inflammatory cytokines also regulate the release of some adipokines and other cytokines, worsening the adipose tissue associated pro-inflammatory state, altering tumor-stroma interactions further activating inflammatory signal pathways thus contributing to the development of invasive PDAC [72].
Pancreas-residing or -infiltrating adipocytes or adipose-like cells such as stromal cells are as important for maintaining pancreatic structure and function as the peripheral and peri-pancreatic adipose tissues. Under obese and pro-tumorigenic conditions, the intra-pancreatic fat-laden cells (steatosis) interact with tumor cells and non-fat stromal or precursor cells, producing an abnormal pro-tumorigenic, pro-inflammatory, and profibrotic symbiosis. With mechanisms similar to the aforementioned for peripheral and peri-pancreatic adipose tissues, this symbiosis promotes the abnormal local release of cytokines, activation of pancreatic stromal cells, infiltration of myofibroblast-like cells and immune cells, accumulation of ECM, angiogenesis, and fibrosis, resulting in pancreatic desmoplasia which, in turn, promotes growth, progression, and treatment resistance of PDAC [22,[73][74][75]. Pancreatic steatosis (fatty pancreas) and the infiltration of the peripancreatic fat are associated with poor clinical outcomes in human PDAC patients [76][77][78]. In KPC (Ptf1 Cre ;KRAS LSL-G12D/+ ;Trp53 LSL-R172H/+ ) mice fed a HFD, the rapid development of invasive PDAC and resistance to chemotherapy were accompanied by a significant increase in the number and size of intra-tumoral fat cells, impaired vascular perfusion, secretion of IL1β, subsequent recruitment of tumor-associated neutrophils, macrophages, lymphocytes, and activation of pancreatic stromal cells [30,68]. In summary, activation of an inflammatory response in the intra-or peri-pancreatic adipose tissues or fat-like cells in the context of obesity induces inflammation and fibrosis in the neighboring pancreas and the cellular precursors to pancreatic cancer, promoting progression to aggressive PDAC and drug resistance.
These observations prompted a number of clinical studies with various inhibitors targeting adipokines, mediators of inflammation and associated secretory factors, and abnormal lipid metabolism for the prevention and treatment of PDAC. However, thus far these inhibitors have shown no notable clinical benefits to PDAC patients. Trials with Cixutumab/Ganitumab, Etanercept, Anakinral, and Tocilizumab that target IGF1R, TNFα, IL1 receptor, and IL6 receptor, respectively, alone or in combination with other chemotherapy agents, have produced disappointing results with no improvements in clinically meaningful parameters such as time to disease progression, median progressionfree survival, disease control rate, or overall survival [79][80][81].
Studies also suggested that broadly targeting the abnormal lipid metabolic profile in PDAC patients could be a viable therapeutic approach. Use of statins that inhibit HMG-CoA reductase mediated cholesterol synthesis has been associated with decreased human pancreatic cancer risk and reduction of pancreatic cancer development and cancer cell growth in pre-clinical studies [82,83]. However, phase 2 randomized clinical trials with simvastatin in advanced PDAC patients showed no improvement in clinical outcomes [84]. Metformin is a widely used treatment agent for type 2 diabetes associated with obesity and the metabolic syndrome thought to function by targeting the mTOR and AMPK pathways. Encouragingly, studies in a pancreatic cancer mouse model with diabetes/obesity showed inhibitory effects of metformin on tumor progression [85]. However, multiple human clinical trials with metformin failed to show advantages in clinical outcomes even in combination with other standard chemotherapies [86][87][88]. In sum, clinical outcomes with inhibitors targeting obesity-associated abnormal lipid metabolism, soluble factors such as cytokines and growth factors, and inflammatory pathways in PDAC patients have so far failed to achieve significant clinical impact despite the promising leads generated by the mechanistic and preclinical insights described above. A deeper understanding of the mechanistic role of mutant KRAS and physiological and biochemical changes precipitated by obesity, diets high in fat, and the associated inflammatory microenvironment is needed to develop novel and effective preventive and therapeutic strategies for PDAC.

Roles of Oncogenic KRAS in the Development of PDAC
The RAS family member genes, including HRAS, NRAS, and KRAS, are the most frequently mutated oncogenes found in as high as 30% of all human cancers, among which mutations in KRAS constitute 86% of all RAS mutations [89,90]. Pancreatic cancerassociated KRAS gene mutations include the missense mutations that result in single amino acid substitutions primarily at codon G12 (98%) with lower frequencies at G13 or Q61. G12D is the predominant mutation accounting for 51% of all mutations at codon G12 (COSMIC, http://cancer.sanger.ac.uk/cancergenome/projects/cosmic/). Despite the critical importance of oncogenic KRAS mutations in PDAC patients, direct molecular targeting of oncogenic KRAS has been thus far problematic. Except for the recent successful development of the specific inhibitors for KRAS G12C , which is found in only 3% of PDAC patients [91][92][93][94][95][96][97], no specific inhibitors for KRAS G12D/V have been reported to date. Thus, new preventive and therapeutic strategies directly or indirectly targeting KRAS G12D/V are critically important for dealing with this devastating disease.
The KRAS gene is a proto-oncogene that encodes a small GTPase, which functions as a binary molecular switch. When bound with guanosine diphosphate (GDP), KRAS is inactivated, which denotes the so-called "off" state, while when bound with guanosine triphosphate (GTP) in response to an external signal or stimulus, KRAS is activated, which signifies a switch to the so-called "on" state [98,99]. The KRAS-GDP to KRAS-GTP conversion is regulated by guanine nucleotide exchange factors (GEFs), such as the Son of Sevenless (SOS), that promote nucleotide exchange and formation of active KRAS-GTP [100]. The activation of KRAS promotes various downstream signaling pathways, such as the MAPK pathway, the PI3K pathway, and the Ral-GEFs pathway, leading to an array of cellular responses, including cell proliferation, growth, survival, migration, and metabolic reprogramming [101,102]. The switch from KRAS-GTP to KRAS-GDP is facilitated by GTPase-activating proteins (GAPs) [103]. KRAS molecules have intrinsic GTPase activity and upon binding with GAPs the rate of GTP hydrolysis is accelerated, generating GDP-bound inactive KRAS, the predominant state under normal physiological conditions [104][105][106]. Thus, the activation of wild-type KRAS is transient ( Figure 1A). conversion is regulated by guanine nucleotide exchange factors (GEFs), such as the Son of Sevenless (SOS), that promote nucleotide exchange and formation of active KRAS-GTP [100]. The activation of KRAS promotes various downstream signaling pathways, such as the MAPK pathway, the PI3K pathway, and the Ral-GEFs pathway, leading to an array of cellular responses, including cell proliferation, growth, survival, migration, and metabolic reprogramming [101,102]. The switch from KRAS-GTP to KRAS-GDP is facilitated by GTPase-activating proteins (GAPs) [103]. KRAS molecules have intrinsic GTPase activity and upon binding with GAPs the rate of GTP hydrolysis is accelerated, generating GDPbound inactive KRAS, the predominant state under normal physiological conditions [104][105][106]. Thus, the activation of wild-type KRAS is transient ( Figure 1A).  Previous studies showed that mutations in KRAS disable its interaction with GAPs, and thus impair its ability to return to the "off" state. For this reason, over the past thirty years, it has been assumed that mutant KRAS is locked in a permanent "on" and constitutively active state, which is capable of driving uncontrolled cell division, growth, survival, invasion, and metabolic maladaptation characteristic of cellular transformation without external stimuli [98,107,108]. Many mouse models have been created to recapitulate the tumorigenic processes driven by oncogenic mutant KRAS, in particular the development of PDAC. The widely used Pdx1-KC (Pdx-1 Cre ;Kras LSL-G12D ) mouse model was generated using a Cre-LoxP technology by crossing the Pdx1 Cre mice with KRAS LSL-G12D/+ mice [109]. Following activation of Cre recombinase, the Pdx1-KC mice simulate an endogenous level of KRAS G12D in pancreatic progenitor cells. The constitutive, endogenous levels of KRAS G12D expression from the embryonic stage induced a full spectrum of pancreatic neoplastic pathologies, including PanIN lesions and PDAC; however, the incidence of PDAC development was only 7% within a year, far less than what was predicted [109]. Substantial evidence supports that acinar cells are the cell of origin for PDAC [110][111][112]. Given that PDAC is likely initiated during adulthood by somatic mutations in KRAS in acinar cells rather than during embryonic development, a mouse model that allowed tetracycline-controlled expression of endogenous levels of KRAS G12V specifically in pancreatic acinar cells in the adult mice was developed [113,114]. This model demonstrated that adult pancreatic acinar cells are rather refractory to transformation by KRAS G12V alone and no PanIN lesions or PDAC foci were observed. Similarly, mouse models expressing oncogenic KRAS at endogenous levels showed that the number of transformed cells was only a small fraction of those expressing oncogenic KRAS [115]. These observations suggest that mutant KRAS alone is insufficient to drive full-blown PDAC and other factors or a second hit is required. This conclusion is supported by studies showing that KRAS mutations are also frequently detected in healthy individuals [116,117].
Comparison of KRAS activity from pancreatic tissue samples expressing one copy of Kras G12D and one copy of wild-type Kras in pancreatic acinar cells to those expressing wildtype KRAS showed that the level of GTP-loaded RAS in cells carrying one copy of Kras G12D was far less than expected if KRAS G12D was constitutively GTP occupied [118][119][120]. Specifically, mutant KRAS activity was less than 2%, which is in marked contrast to what was expected when 50% of the total KRAS protein was mutated [120]. Furthermore, the expression of endogenous levels of mutant KRAS failed to significantly activate its downstream signaling pathways and to induce notable pathological changes [113,115]. Direct evaluation of GTP occupation of KRAS indicated that stimulants such as EGF could significantly increase GTP-bound active KRAS for a prolonged time in cells bearing endogenous levels of mutant KRAS [115]. Expression of a dominant-negative GEF reduced RAS activity in cells expressing mutant RAS [121], suggesting that, rather than being constitutively active, mutant RAS activity is still subject to the regulations by GEF as seen with the wild-type KRAS. Upon interaction with GEFs, mutant KRAS exhibited prolonged activation relative to wild-type KRAS [120]. These observations suggest that mutant KRAS is not fully active as previously thought and can be further enhanced by upstream stimuli with prolonged activation. Overexpression of mutant KRAS by strong promoters was highly effective in transforming cells [122], which was attributed to the high levels of KRAS activity. This is in contrast to endogenous levels of mutant KRAS, which appears to exhibit lower levels of KRAS activity and is ineffective in transforming cells ( Figure 1B). Unlike what had been assumed to be the pervasive roles of mutant KRAS in pancreatic oncogenesis, these data suggest that an endogenous level of oncogenic KRAS is far less than fully active and has limited capacity to drive the development of PDAC and therefore, a second hit is required for enhancing oncogenic KRAS activity that can drive full-blown PDAC. In this regard, obesity and inflammation are considered among the most predominant clinical risk factors and are likely to synergize with mutant KRAS to drive pancreatic tumorigenesis leading to invasive PDAC, in which oncogenic KRAS is known to be hyperactivated ( Figure 1C).
These studies suggest that the regulation of mutant KRAS activity is complex and our understanding of these mechanisms is currently incomplete.

Interplay between Obesogenic HFD, Inflammation, and Oncogenic KRAS in PDAC Development
Although several genetically engineered mouse models mimicking endogenous levels of mutant KRAS expression in pancreatic acinar cells demonstrated that oncogenic KRAS alone is marginally effective at inducing full-blown PDAC, the combination of oncogenic KRAS and chronic HFD consumption led to extensive pancreatic inflammation, fibrosis, high-grade PanIN lesions, and invasive PDAC [49][50][51][52][53]. In addition, KRAS G12D mice fed a HFD exhibited significantly increased levels of glycolytic enzymes including hexokinase II and lactate dehydrogenase A compared to mice fed a control diet, indicating that a key metabolic effect of mutant KRAS (e.g., enhancement of anaerobic glycolysis) is magnified in the setting of obesogenic HFD challenge [123]. These observations indicate that mutant KRAS renders mice more susceptible to adverse insults of chronic HFD, leading to the enhancement of oncogenic KRAS-mediated development of invasive PDAC.
Like obesity, chronic pancreatitis is also a known risk factor for PDAC development. When subjected to caerulein (an agonist of cholecystokinin (CCK))-induced pancreatitis, adult mice that express endogenous levels of KRAS G12V in the pancreatic acinar cell compartment developed extensive PanIN lesions and PDAC with high penetrance [113,114]. Similarly, overexpression of cyclooxygenase-2 (COX-2) in pancreatic acinar cells of fElas CreERT -KC mice that express endogenous levels of KRAS G12D induced RAS hyperactivation and associated extensive inflammation, fibrosis, and rapid development of high-grade PanIN lesions [115]. Consistent with these observations, constitutive activation of NFκB by overexpressing IKK2 (inhibitor kappa B kinase 2) in pancreatic acinar cells of fElas CreERT -KC mice led to the rapid development of neoplastic lesions [115]. These results are also in line with other studies showing that although the insulin-expressing endocrine cells were refractory to oncogenic KRAS-induced transformation, chronic pancreatic inflammation could cooperate with oncogenic KRAS to alter the endocrine phenotype of these cells, leading to pancreatic neoplastic progression in cells generally refractory to neoplastic transformation [124]. These experimental studies highlight the critical role of inflammation in promoting oncogenic KRAS-mediated development of pancreatic neoplastic lesions, supporting an important synergy between mutant KRAS and inflammation in driving PDAC development [113,114,[124][125][126]. Furthermore, inhibition of inflammation by ablating COX-2 or IKK2 curbed the synergy between inflammation and oncogenic KRAS, leading to significant suppression of inflammation and fibrosis, PanIN lesions, and PDAC development [115].
Notably, studies showed that the inflamed pancreata from the aforementioned mice suffering from obesity and extensive inflammation exhibited a significant increase in KRAS activity ( Figure 1C). In addition, when treated with inflammatory stimulants, such as CCK, prostaglandin E2, or lipopolysaccharide, the amount of GTP-bound mutant KRAS is increased and prolonged compared to that of wild-type KRAS [115]. In marked contrast, inhibition of HFD challenge by the administration of recombinant FGF21 (see below) or suppression of inflammatory stimuli by ablation of COX-2 or IKK2 curbed RAS activity with notable attenuation of PanIN lesions and malignant progression to PDAC [49,53,115,123]. These studies suggest that the hyperactivation of mutant KRAS by obesogenic HFD and inflammation, rather than the simple presence of an endogenous level of mutant KRAS protein alone, is critical for the development of PDAC [118,119]. Although oncogenic mutation in KRAS is a necessary genetic transforming event and confers susceptibility to additional challenges, obesogenic HFD challenge or inflammatory insults serve as the second hit to enhance KRAS oncogenic capacity, driving pancreatic tumorigenesis and the eventual development of invasive PDAC. Mechanistically, the prolonged activation of mutant KRAS led to the activation of NFκB, COX-2, and others, creating a positive feedback loop to sustain mutant KRAS hyperactivity, leading to the rapid development of invasive PDAC [49,115]. This appears to be true in the early stage of PDAC development since the ablation of COX-2 or IKK2, as well as inhibition of COX-2 by celecoxib, suppressed mutant KRAS-mediated PDAC development [49,115]. However, interestingly, in vitro studies showed that neither the presence of oncogenic KRAS nor the level of mutant KRAS activity positively correlated with COX-2 protein levels in pancreatic cancer cells [120,127,128], suggesting that COX-2 overexpression and resultant inflammation might be a critical early event in KRAS hyperactivation during PDAC development. When the activity of mutant KRAS reaches a certain level or a threshold, external stimuli, such as COX-2, may no longer be required for the development of the late-stage PDAC. This may in part explain the failure of the inhibitors of obesogenic or inflammatory factors as aforementioned in deterring PDAC development in patients even in combination with other chemotherapy agents that do not target mutant KRAS. Recently, it has also been shown that there are two distinct inflammation states associated with the macrophages and Th cells in pancreatic cell microenvironment during PDAC development, the so-called Th1-M1 state and the Th2-M2 state, which correspond to an inflammatory and an anti-inflammatory phenotype, respectively. Evidence suggests that the Th1-M1 state may predominate the early ADM and PanIN stages, while the PDAC is mostly associated with the anti-inflammatory Th2-M2 state. This may also explain why anti-inflammatory agents have little effect on diminishing PDAC progression while possibly having a role in preventing the precursory ADM and early PanIN lesions [72,129,130]. Therefore, a detailed understanding of the vulnerability of the pancreas to exogenous risk factors during PDAC development and the underlying mechanisms of the interplay between oncogenic KRAS and the environmental risk factors, including obesity, chronic HFD, and inflammation, are critical for the design of novel, effective targeting strategies for the prevention and therapy of pancreatic cancer.

FGF21: A Potential Missing Link between Obesity/Inflammation and Mutant KRAS-mediated Pancreatic Tumorigenesis
As a member of the FGF family, FGF21 is unique in that it functions to maintain lipid and energy homeostasis under normal physiological conditions, promoting metabolic homeostasis rather than stimulating cell proliferation and growth as most FGF family members do. Pharmacological levels of FGF21 exert both anti-obesity and anti-inflammatory activities. As an endocrine hormone, FGF21 is produced in the liver in response to local liver perturbations as well as systemic metabolic derangements and acts primarily on white and brown adipose tissues and the hypothalamus where the canonical transmembrane fibroblast growth factor receptor 1 (FGFR1) kinase and a newly described accessory nonkinase membrane-bound co-receptor Klotho beta (KLB) complex is located to promote the correction of the deranged metabolic parameters and the maintenance of metabolic homeostasis. The serum levels of FGF21 are increased in the settings of hyperlipidemia, obesity, diabetes, and fatty liver [131][132][133]. Adipose tissues also produce FGF21 as an adipokine, which potentially acts in a paracrine and autocrine mode [134][135][136]. Administration of exogenous FGF21, FGF21 analogs, or FGFR1-KLB agonists promotes an array of metabolic benefits in mice that intercept obesity, type 2 diabetes, fatty liver disease, and hyperlipidemia [137][138][139][140][141]. It is important to note that adipose tissue is the primary as well as the ultimate (in terms of some of the central nervous system routes) target for major effects of FGF21 where its receptor complex FGFR1-KLB is predominantly located [138,[142][143][144][145]. The pharmacologic impact of FGF21 on obese and inflamed adipose tissues, residential adipocytes, and the associated metabolic abnormalities and inflammatory damage in the context of obesity is expected to be more direct and significant than some of the known metabolic and inflammatory regulators, such as leptin, adiponectin, and IL6.
A number of studies have revealed that FGF21 is expressed also at high levels at normal conditions in the pancreas, in particular the pancreatic acinar cells, which can be further induced in response to pancreatic perturbations, such as the fatty pancreas, pancreatitis, and pancreatic injury [53,146,147]. FGF21 also functions as a secretagogue to promote pancreatic exocrine function and maintain acinar cell proteostasis, which if uncontrolled could cause pancreatitis. Loss of FGF21 results in abnormal accumulation of zymogen granules that induces pancreatic ER stress (Table 1) [148]. Similarly, mice lacking FGF21 developed significant islet hyperplasia and periductal lymphocytic inflammation with elevated expression of cytokines such as TNFα, interferon gamma, and IL1β when fed an obesogenic diet [147], while liver-specific overexpression of FGF21 reduces pancreatic β-cell apoptosis in db/db mice (Table 1) [149]. Consistent with the beneficial effects of FGF21 on obesity and fatty liver disease, pharmacological FGF21 protects the pancreas from gluco-lipotoxicity and cytokine-induced damage, pancreatitis, and fibrosis both in the islet and acinar cell compartments, promoting both endocrine and exocrine functions of the pancreas in various mouse models, such as the Streptozotocin (STZ)-induced type 1 diabetes, ob/ob, db/db, and diet-induced obese (DIO) models [150][151][152][153][154][155]. Deficiency of FGF21 under pancreatitis induced by caerulein, thapsigargin, or mechanical stress worsened pancreatitis and precipitated overt organ damage, while overexpression of FGF21 reversed these effects (Table 1) [156]. FGF21 treatment in various mouse models of acute and chronic pancreatitis induced by caerulein, alcohol consumption, or endoscopic retrograde cholangiopancreatography radiocontrast dye prevented the onset of pancreatitis and alleviated symptoms, while acinar cell-specific deficiency or silencing of FGF21 or its co-receptor KLB blocked these beneficial effects [156,157]. These data suggest that FGF21 deficiency is necessary for pancreatitis and that FGF21 has the potential for treating pancreatitis. Table 1. The effects of fibroblast growth factor 21 (FGF21) signaling in the pancreas.

Streptozotocin
[STZ]-induced type 1 diabetes Improves islet engraftment and insulin sensitivity. [154] Prevents the increase in glycemia and lowers lipids. [159] Injection of FGF21 ob/ob and DIO Improves glucose tolerance and insulin sensitivity, but has no direct effect on islet insulin secretion. [155] Injection of FGF21 or analogs db/db Improves islet survival and function, insulin sensitivity and glucose homeostasis. [150,160] Increases pancreatic β-cell mass [161] Both the pancreas and the liver are metabolic organs and are subject to the development of a spectrum of the fatty pancreas and fatty liver diseases under conditions of chronic caloric excess. The protective anti-obesity (including the obesity complications), anti-diabetes, and anti-inflammation effects of pharmacological FGF21 have been extensively studied in models of fatty liver disease including hepatosteatosis, non-alcoholic steatohepatitis (NASH), and in malignancies such as hepatocellular carcinoma [162][163][164][165][166][167].
In clinical trials with patients who were obese or diabetic with a predisposition to fatty liver complications including NASH, long-acting FGF21 analogs improved serum and hepatic metabolic parameters and hepatic fibrosis markers, suggesting that FGF21 is a promising therapy for obesity-and meta-inflammation-associated liver diseases [140,168,169]. Based on the resemblance in the pathological spectrums between fatty liver disease and fatty pancreas disease, it can be extrapolated that FGF21 would similarly exert an inhibitory, protective effect on the fatty pancreas and inflammation-associated PDAC development.
Furthermore, in several mouse models of mutant KRAS G12D/+ -mediated PDAC as well as in patient samples, the expression of FGF21 is negatively associated with the expression of mutant KRAS G12D/+ , suggesting that FGF21 is potentially a downstream target of oncogenic KRAS [53]. Silencing the expression of FGF21 by oncogenic KRAS is proposed to create a vulnerability in acinar cells to metabolic and inflammatory challenges, leading to a loss of a cell-autonomous defense system against metabolic and inflammatory insults that in concert with oncogenic KRAS promotes neoplastic transformation and accelerates the transition of low-grade PanIN lesions to high-grade lesions and invasive PDAC ( Figure 2). Administration of pharmacological FGF21 compensates for the loss of pancreatic FGF21 induced by mutant KRAS G12D in the context of an obesogenic HFD, leading to significant suppression of PanIN lesions, inhibition of invasive PDAC development, and reduction in liver metastasis [53]. Positive physiologic effects were also observed including inhibition of weight gain, pancreatic local and systemic inflammation, and pancreatic fibrosis. Taken together, FGF21 functions, at least in part, to offset the vulnerability that is incurred by oncogenic KRAS and hijacked by obesity and other inflammatory states. FGF21 may represent a promising next-generation preventive and therapeutic strategy that warrants comprehensive testing in human clinical trials of obesity-associated pancreatic cancer, pancreatitis, and other inflammation-related disease states such as NASH and hepatocellular cancer.
Both the pancreas and the liver are metabolic organs and are subject to the development of a spectrum of the fatty pancreas and fatty liver diseases under conditions of chronic caloric excess. The protective anti-obesity (including the obesity complications), anti-diabetes, and anti-inflammation effects of pharmacological FGF21 have been extensively studied in models of fatty liver disease including hepatosteatosis, non-alcoholic steatohepatitis (NASH), and in malignancies such as hepatocellular carcinoma [162][163][164][165][166][167]. In clinical trials with patients who were obese or diabetic with a predisposition to fatty liver complications including NASH, long-acting FGF21 analogs improved serum and hepatic metabolic parameters and hepatic fibrosis markers, suggesting that FGF21 is a promising therapy for obesity-and meta-inflammation-associated liver diseases [140,168,169]. Based on the resemblance in the pathological spectrums between fatty liver disease and fatty pancreas disease, it can be extrapolated that FGF21 would similarly exert an inhibitory, protective effect on the fatty pancreas and inflammation-associated PDAC development.
Furthermore, in several mouse models of mutant KRAS G12D/+ -mediated PDAC as well as in patient samples, the expression of FGF21 is negatively associated with the expression of mutant KRAS G12D/+ , suggesting that FGF21 is potentially a downstream target of oncogenic KRAS [53]. Silencing the expression of FGF21 by oncogenic KRAS is proposed to create a vulnerability in acinar cells to metabolic and inflammatory challenges, leading to a loss of a cell-autonomous defense system against metabolic and inflammatory insults that in concert with oncogenic KRAS promotes neoplastic transformation and accelerates the transition of low-grade PanIN lesions to high-grade lesions and invasive PDAC (Figure 2). Administration of pharmacological FGF21 compensates for the loss of pancreatic FGF21 induced by mutant KRAS G12D in the context of an obesogenic HFD, leading to significant suppression of PanIN lesions, inhibition of invasive PDAC development, and reduction in liver metastasis [53]. Positive physiologic effects were also observed including inhibition of weight gain, pancreatic local and systemic inflammation, and pancreatic fibrosis. Taken together, FGF21 functions, at least in part, to offset the vulnerability that is incurred by oncogenic KRAS and hijacked by obesity and other inflammatory states. FGF21 may represent a promising next-generation preventive and therapeutic strategy that warrants comprehensive testing in human clinical trials of obesity-associated pancreatic cancer, pancreatitis, and other inflammation-related disease states such as NASH and hepatocellular cancer. The mechanistic interplay of mutant KRAS and obesity/inflammation. Oncogenic KRAS not only directly or indirectly promotes cell proliferation, resistance to apoptosis, alteration of metabolism but also silences the expression of FGF21, a novel anti-obesity and anti-inflammation factor, leading to irreversible acinar-to-ductal metaplasia (ADM), fibrosis, and PanIN lesions. The loss of acinar cell FGF21 (and the alteration of other genes) by mutant KRAS promotes vulnerability in the acinar cell compartment to chronic HFD or inflammatory challenge, leading to KRAS hyperactivation, accentuated pancreatic inflammation, advanced PanIN lesions, and subsequent development of PDAC with high penetrance. Images were from the experiments of our laboratory. Figure 2. The mechanistic interplay of mutant KRAS and obesity/inflammation. Oncogenic KRAS not only directly or indirectly promotes cell proliferation, resistance to apoptosis, alteration of metabolism but also silences the expression of FGF21, a novel anti-obesity and anti-inflammation factor, leading to irreversible acinar-to-ductal metaplasia (ADM), fibrosis, and PanIN lesions. The loss of acinar cell FGF21 (and the alteration of other genes) by mutant KRAS promotes vulnerability in the acinar cell compartment to chronic HFD or inflammatory challenge, leading to KRAS hyperactivation, accentuated pancreatic inflammation, advanced PanIN lesions, and subsequent development of PDAC with high penetrance. Images were from the experiments of our laboratory.

Concluding Remarks and Future Perspectives
PDAC is currently one of the deadliest cancers in part owing to the lack of effective preventive and therapeutic approaches. Point mutations in the KRAS proto-oncogene, in particular G12D and G12V, are the dominant drivers of the initiation and progression of PDAC. Yet, inhibitors specifically targeting the major KRAS oncoproteins relevant to pancreatic cancer have not been successfully developed. Of note, mounting evidence suggests that mutant KRAS alone is insufficient to drive PDAC development in adults and environmental risk factors such as obesity and pancreatic inflammation are found to synergize with mutant KRAS to induce oncogenic KRAS hyperactivation, promoting extensive inflammation, fibrosis, and the rapid development of PanIN lesions and PDAC with high penetrance. The recent discovery of the potent metabolic regulator FGF21 with a combined activity of anti-obesity and anti-inflammation and as a downstream target of oncogenic KRAS has raised new optimism for PDAC prevention and therapy. FGF21 analogs have been tested in obese or diabetic patients with NASH generating favorable outcomes and exhibiting inhibitory effects on hepatocellular carcinoma in association with fatty liver disease in animal models. Notably, pharmacological FGF21 has also been shown to inhibit not only KRAS G12D -mediated pancreatic inflammation, fibrosis, and PanIN lesions but also the development of malignant PDAC and liver metastasis even in the setting of obesogenic HFD challenge [53]. Future work should focus on further delineating the mechanisms of KRAS hyperactivation by obesity, chronic HFD, and inflammation, the critical factors required to sustain the prolonged KRAS activity in PDAC development, and how FGF21 analogs or pathway agonists should be utilized in a clinical setting for pancreatic cancer patients who also suffer from the effects of longstanding obesity and/or pancreatitis.