CXCL10/CXCR3 signaling contributes to an inflammatory microenvironment and its blockade enhances progression of murine pancreatic precancerous lesions
Abstract
The development of pancreatic cancer requires recruitment and activation of different macrophage populations. However, little is known about how macrophages are attracted to the pancreas after injury or an oncogenic event, and how they crosstalk with lesion cells or other cells of the lesion microenvironment. Here, we delineate the importance of CXCL10/CXCR3 signaling during the early phase of murine pancreatic cancer. We show that CXCL10 is produced by pancreatic precancerous lesion cells in response to IFNγ signaling and that inflammatory macrophages are recipients for this chemokine. CXCL10/CXCR3 signaling in macrophages mediates their chemoattraction to the pancreas, enhances their proliferation, and maintains their inflammatory identity. Blocking of CXCL10/CXCR3 signaling in vivo shifts macrophage populations to a tumor-promoting (Ym1+, Fizz+, Arg1+) phenotype, increases fibrosis, and mediates progression of lesions, highlighting the importance of this pathway in PDA development. This is reversed when CXCL10 is overexpressed in PanIN cells.
Introduction
Pancreatic cancer is difficult to target because its fibrotic microenvironment not only acts as a barrier for delivery of tumor cell targeting drugs, but it also generates an anti-inflammatory environment and prevents immunotherapy (Balachandran et al., 2019). One of the current paradigms for treatment of PDA focuses on combining chemotherapy with immune modulators that reprogram tumor-promoting macrophages toward a pro-inflammatory phenotype (Bastea et al., 2019; Mitchem et al., 2013; Pandey and Storz, 2019). A deeper understanding of the mechanisms that play a role in macrophage polarization can provide insights to develop such new interventions.
Genetic mouse models have shown that pancreatic ductal adenocarcinoma (PDA) most likely originates from precancerous pancreatic intraepithelial neoplasm (PanIN) lesions (reviewed in Storz, 2017). The development and progression of these early lesions is dependent on crosstalk between a multitude of host cells in their microenvironment. Of these, inflammatory (M1-polarized) and alternatively activated (M2-polarized) macrophages are the most consequential cell types.
The initial influx of macrophages, which induces local inflammation, occurs in response to an aberrant release of chemokines from pancreatic cells undergoing transformation (Liou et al., 2015). However, local inflammation alone is not an efficient driver of oncogenic progression and requires additional inflammatory signaling, genetic alterations, and downregulation of factors that maintain acinar cell identity (Carrière et al., 2011; Cobo et al., 2018; Guerra et al., 2011; Guerra et al., 2007).
Inflammatory macrophages contribute to pre-neoplastic lesion formation via secretion of inflammatory mediators, which regulate reorganization of the acinar microenvironment and initiate acinar-to-ductal metaplasia (ADM) (Liou et al., 2015; Liou et al., 2013; Sawey et al., 2007). While pro-inflammatory M1 macrophages are important for the initiation of precancerous lesions, this population dwindles and M2 macrophages become more predominant (Liou et al., 2017). These M2 macrophages are chitinase-like protein 3 (Ym1/Chil3), arginase-1 (Arg1), resistin-like alpha (Fizz1/Retnla), and interleukin-1 receptor antagonist protein (IL-1ra) positive and promote lesion growth, drive fibrogenesis, and block T-cell infiltration (Bastea et al., 2019; Liou et al., 2017). Later, at the tumor stage, alternatively activated macrophages represent approximately 85% of tumor-associated macrophages (TAMs) in the microenvironment (Partecke et al., 2013). For full-blown pancreatic cancer, tissue-resident macrophages have been suggested to shape fibrotic responses (Zhu et al., 2017), while infiltrating monocytes generate an immunosuppressive environment (Zhang et al., 2017; Zhu et al., 2017).
C-X-C motif chemokine 10 (CXCL10), also known as IFNγ-induced protein 10 (IP-10), acts through its cognate receptor C-X-C motif chemokine receptor 3 (CXCR3) (Groom and Luster, 2011) and regulates the chemotaxis of CXCR3+ immune cells such as macrophages, T cells, and natural killer (NK) cells (Luster and Ravetch, 1987; Tomita et al., 2016; Zhou et al., 2010). With respect to cancer aggressiveness and patient prognosis, the presence of CXCL10 and CXCR3 has shown conflicting results depending on the type and stage of the disease (Fulton, 2009; Jacquelot et al., 2018; Li et al., 2015). In pancreatic cancer, both CXCL10 and CXCR3 are expressed in tumor tissue (Delitto et al., 2015), and their presence has been correlated with poor prognosis (Liu et al., 2011; Lunardi et al., 2014). However, the role of CXCL10/CXCR3 signaling during early development of the disease has not been addressed.
In our present study, we show that CXCL10, produced by precancerous lesions cells, is involved in the onset of inflammation by chemoattracting macrophages. We further show that CXCL10 signaling to CXCR3 is a key event for inflammatory macrophage identity and that inhibition of CXCL10/CXCR3 signaling leads to a polarization shift to an alternatively activated phenotype. In vivo, we demonstrate the importance of CXCL10/CXCR3 signaling in the maintenance of an inflammatory microenvironment, and that its blockage drives tumor progression.
Results
Pre-neoplastic ADM and PanIN lesions produce CXCL10
To identify factors that are released by precancerous lesion cells, we performed a cytokine/chemokine assay. Therefore, we used SM3 cells, which have been isolated from the precancerous epithelium of a KC mouse and form lesions with PanIN features when cultivated on extracellular matrix (Agbunag et al., 2006; Liou et al., 2017). In this screen limited to these in vitro lesion cells (Figure 1—figure supplement 1A), besides known factors such as C-C motif chemokine 5 (CCL5) and metalloproteinase inhibitor 1 (TIMP-1), we found strong expression of CXCL10, which has previously been identified as a chemoattractant for macrophages (Tomita et al., 2016; Zhou et al., 2010). We then used fluorescent in situ hybridization (FISH) to determine whether Cxcl10 is produced in pancreatic precancerous lesion areas of Ptf1a/p48cre;LSL-KrasG12D (KC) mice. While Cxcl10 was undetectable in normal adjacent acini (Figure 1—figure supplement 1B), we found significant expression of Cxcl10 in ADM and PanIN1 lesions (Figure 1A). Quantification analyses of samples stained for Cxcl10 mRNA by ISH indicated approximately fivefold higher expression in ADM than in PanIN (Figure 1B, Figure 1—figure supplement 1C). Next, we isolated primary pancreatic acinar cells from LSL-KrasG12D mice and adenovirally infected them with either GFP (control) or Cre-GFP, to test whether Cxcl10 expression is upregulated during the KRasG12D-driven ADM process (Figure 1—figure supplement 1D). However, expression of KRasG12D was unable to increase Cxcl10 expression, indicating an external stimulus as a driver.
CXCL10 (also IP-10, interferon gamma-inducible protein 10) expression has previously been shown to be induced by interferon gamma (IFNγ) via activation of signal transducer and activator of transcription 1 (STAT1) (Han et al., 2010; Luster and Ravetch, 1987). Therefore, we tested if this pathway is active in PanIN cells. Treatment of SM3 cells with IFNγ induced an over 60-fold increase in Cxcl10 mRNA (Figure 1C), as well as increased CXCL10 protein production (Figure 1D) and secretion (Figure 1E). To test whether CXCL10 expression is indeed mediated through STAT1 signaling, we combined IFNγ stimulation with the pan-JAK inhibitor NVP-BSK805. We found that IFNγ stimulation led to phosphorylation of STAT1 at Y701 (activating phosphorylation), increased expression of CXCL10, and that pre-treatment with NVP-BSK805 inhibited IFNγ-induced pY701-STAT1 and CXCL10 expression (Figure 1F). T cells and NK cells are known IFNγ producers in the pancreatic microenvironment (Brauner et al., 2010; Chapoval et al., 2001; Loos et al., 2009). To determine whether these cells could be an in vivo source for IFNγ in our mouse model, we performed an ISH for Ifng combined with IHC for T-cell surface glycoprotein CD4 (CD4), T-cell surface glycoprotein CD8 (CD8), or NKG2-D type II integral membrane protein (NKG2D) markers. As expected from published data, we found both T cells and NK cells as a potential source for IFNγ (Figure 1G).
Inflammatory macrophages are the recipients for CXCL10
With respect to early events leading to development of PDA, the influx of macrophages into the pancreas has been demonstrated following injury and during development and progression of pancreatic lesions (Gea-Sorlí and Closa, 2009; Liou et al., 2015). Moreover, CXCL10 has been demonstrated as a chemoattractant for macrophages along with other immune cells (Liu et al., 2011). This prompted us to test whether macrophages are responsive to CXCL10. We found that non-polarized peritoneal macrophages express high levels of the CXCL10 receptor Cxcr3, while M1-polarized (inflammatory) macrophages express moderate levels, and M2-polarized (alternatively activated) macrophages do not express this receptor (Figure 2A). Transwell invasion assays using both peritoneal and bone marrow-derived macrophages suggest that CXCL10 can act as a chemoattractant for M1-polarized macrophages (Figure 2B, Figure 2—figure supplement 1A). However, since tissue-resident macrophages have been attributed important roles in established pancreatic cancer (Zhu et al., 2017), we also determined if this population can be the recipients for CXCL10. Approximately 80% of tissue resident macrophages in normal mouse pancreas express CXCR3 (Figure 2—figure supplement 1A), but when isolated, these cells do not proliferate in response to CXCL10 (Figure 2—figure supplement 1B). In sum, our in vitro data suggests that CXCL10 may drive the chemoattraction of inflammatory macrophages to the pancreas.
Next, we determined if pancreatic macrophages or T cells express CXCR3 in KC mice. Therefore, we sorted for pancreatic CD3+ or F4/80+ cells and then for the presence of CXCR3. We found that approximately 40% of pancreatic macrophages in KC mice express CXCR3 (Figure 2C). Moreover, an in situ IF-IHC analysis of pancreata of KC mice indicated that inflammatory (F4/80+;pY701-STAT1+) macrophages express CXCR3, while alternatively activated (F4/80+;Ym1+) macrophages do not express this receptor (Figure 2D), which also confirmed above in vitro data. An overlay between an ISH for Cxcr3 and IF-IHC for inflammatory macrophages (CD68+;iNOS+) in human patient tumors showed that ~70% of CXCR3+ cells are M1 macrophages and confirmed this population as a potential recipient for CXCL10 (Figure 2E,F).
CXCL10/CXCR3 signaling maintains the inflammatory phenotype of macrophages
Next, we tested the impact of blocking CXCL10/CXCR3 signaling on the inflammatory macrophage population. We isolated peritoneal primary macrophages, polarized them to inflammatory (F4/80+;iNOS+) macrophages (Figure 3—figure supplement 1A), and then blocked CXCL10/CXCR3 signaling with a CXCR3 neutralizing antibody (CXCR3 NAB). We found that upon CXCR3 neutralization, M1 macrophages, although still iNOS positive, start to express Ym1 (Figure 3A), which we previously have identified as a bona fide marker for M2 alternatively activated macrophages in precancerous abnormal pancreas lesions (Bastea et al., 2019). A more thorough analysis using qRT-PCR indicated a significant decrease in CD38, which is a distinct M1 macrophage marker (Jablonski et al., 2015), and an upregulation of Ym1, Arg1 and Fizz1 (Figure 3B), which are all markers of the alternatively activated macrophage population that previously has been shown to drive fibrosis during pancreatic lesion progression (Liou et al., 2017). CXCR3 neutralization in M1 macrophages also slightly increased Il4ra and Stat6 mRNA expression, and increased nuclear localization of active Y641-phosphorylated STAT6 (Figure 3—figure supplement 1B,C), which are additional indicators of M2 polarization (Yushi et al., 2016).
The interferon regulatory factors (IRFs) IRF4 and IRF5 have been shown to be key regulators of macrophage polarization (Bastea et al., 2019; Günthner and Anders, 2013). M1-polarized macrophages subjected to CXCR3 inhibition upregulated IRF4, a transcription factor for M2 macrophage polarization, and downregulated the M1 transcription factor IRF5 (Figure 3—figure supplement 1D), indicating a potential mechanism of how pancreatic macrophage populations may shift when CXCR3 is blocked.
To test whether neutralization of CXCL10/CXCR3 signaling also decreases inflammatory macrophages in vivo, we treated KC mice with a CXCR3 neutralization antibody (CXCR3 NAB) or an isotype control IgG antibody over a period of 9 weeks (Figure 3—figure supplement 1E). While overall numbers of macrophages remained unchanged (Figure 3—figure supplement 1F), IF-IHC analyses for F4/80+;pY701-STAT1+ cells indicated a >50% decrease in inflammatory macrophages in the CXCR3 NAB-treated mice (Figure 3C,D).
Neutralization of CXCL10/CXCR3 signaling increases alternatively activated pancreatic macrophages, SMA-positive fibroblasts, and areas of pancreatic lesions
We then determined if the decrease in inflammatory macrophages after neutralization of CXCL10/CXCR3 signaling can lead to an increase in Ym1+ alternatively activated macrophages. The presence of Ym1+ macrophages in the pancreas was increased by approximately threefold when KC mice were treated with a CXCR3 NAB, as compared to treatment with the isotype control IgG (Figure 4A,B). Similar effects on this macrophage population were observed when KC mice were treated with a CXCL10 NAB, as compared to treatment with the isotype control IgG (Figure 4—figure supplement 1A–C).
Previously, Ym1+ macrophages in the KC animal model have been shown to drive fibrosis during pancreatic cancer development (Liou et al., 2017), and after neutralization of CXCR3, or alternatively neutralization of CXCL10, we observed a correlating increase in fibrosis (Figure 4C,D and Figure 4—figure supplement 1D,E). In line with increased presence of alternatively activated macrophages and increased fibrosis, abnormal areas (lesions and stroma) increased approximately twofold when KC mice were treated with CXCR3 NAB (Figure 4E,F). Of note, neutralization of CXCR3 in non-transgenic (ntg) mice did not show any change in the normal pancreatic tissue (Figure 4E). Quantification of different lesion types from both treatment groups indicated slight, but not statistically significant differences in the distribution of ADM (80.7 ± 3.8% in control-treated versus 75.2 ± 4% CXCR3 NAB-treated) and PanIN1 (19.3 ± 3.8% in control-treated versus 24.6 ± 4% CXCR3 NAB-treated) lesions (Figure 4G).
Neutralization of CXCL10/CXCR3 signaling does not significantly affect the presence of T cells
In addition to macrophages, T cells can be chemoattracted via CXCR3 signaling (Dufour et al., 2002). Since approximately 15% of CD3+ T cells express CXCR3 in our KC animal model (see Figure 2C), we also tested the effect of CXCL10/CXCR3 neutralization on T-cell populations in pancreata of KC mice. After neutralization of CXCR3, we detected a slight (approximately 20%) decrease in pancreatic CD3+ cells (Figure 4—figure supplement 2A), but did not observe any changes in NK (NKG2D+;CD3−) cells (Figure 4—figure supplement 2B) or production of IFNγ (Figure 4—figure supplement 2C). Moreover, neutralization of CXCL10 did not result in a decrease of CD3+;CD4+ or CD3+;CD8+ T cells (Figure 4—figure supplement 2D,E).
T cells do not contribute to CXCL10/CXCR3-mediated effects on macrophage populations or abnormal pancreatic lesions
To test whether effects on macrophage populations or abnormal pancreatic lesion areas are due to the presence of T cells, we performed a T-cell depletion followed by CXCR3 neutralization (Figure 4—figure supplement 3A). T cells in KC mice were depleted using previously described antibodies that target CD8α and CD4 (Laky and Kruisbeek, 2016). Regardless of T cells being depleted or present (Figure 4—figure supplement 3B), CXCR3 neutralization increased abnormal pancreatic area (Figure 4—figure supplement 3C), decreased M1 macrophages (Figure 4—figure supplement 3D), and increased Ym1+ macrophages (Figure 4—figure supplement 3E). Moreover, T-cell depletion led to a slight but non-significant decrease in Ifng, suggesting that T cells are not the only source for IFNγ in this experimental system (Figure 4—figure supplement 3F).
To rigorously test a direct effect of CXCL10/CXCR3 signaling on macrophages in a system absent of T cells, we expressed CXCL10 in SM3 PanIN cells and implanted them into the pancreas of T cell-deficient athymic nude mice (Figure 4—figure supplement 3G). Therefore, we infected SM3 PanIN cells (described in Liou et al., 2017) with a lentivirus carrying CXCL10 (SM3-CXCL10) or a control lentivirus carrying eGFP (SM3-control), tested them for secretion of CXCL10 (Figure 4—figure supplement 3H), and generated PanIN organoids by plating them in Matrigel for 2 days (Figure 4—figure supplement 3I). Organoids were recovered from Matrigel and, together with activated stellate cells, implanted into the pancreas of nude mice. Two weeks after implantation, SM3-CXCL10 implanted mice had approximately half as much abnormal tissue area (lesions and fibrosis) as compared to SM3-control implanted mice (Figure 5A,B). Analysis of SMA content showed significantly reduced fibrosis in the SM3-CXCL10 implanted mice (Figure 5C). As expected from the reverse experiment in which we had blocked CXCR3 (Figure 3D), expression of CXCL10 increased the presence of F4/80+ macrophages (Figure 5D). Further analysis of macrophage polarization types indicated an increase of M1 and a decrease of M2 phenotypes when CXCL10 was present (Figure 5E).
Discussion
While inflammatory and alternatively activated macrophages have important functions in different stages of pancreatic cancer development (Stone and Beatty, 2019; Storz, 2017), little is known about how these populations are regulated after injury or an oncogenic event, and how they crosstalk with lesion cells or other cells of the lesion microenvironment. In this study, we delineate the importance of CXCL10/CXCR3 signaling during the early phase of pancreatic cancer development.
The CXCL10/CXCR3 signaling axis has been implicated in generating inflammation in various diseases including pancreatitis (Lee et al., 2017; Liu et al., 2011; Singh et al., 2007). This is because it regulates the chemotaxis of CXCR3+ immune cells such as macrophages, T cells, and NK cells (Luster and Ravetch, 1987; Taub et al., 1993; Tomita et al., 2016; Zhou et al., 2010). With respect to cancer, gastric cancer patients with upregulated CXCR3 expression showed better survival (Chen et al., 2019; Li et al., 2015). Also, high expression of CXCL10 and other CXCR3 ligands in ovarian, esophageal, and non-small cell lung carcinoma indicate favorable prognoses (Bronger et al., 2016; Cao et al., 2017; Sato et al., 2016). In contrast, in pancreatic cancer, expression of both CXCL10 and CXCR3 in tumor tissue has been correlated with a poor prognosis (Liu et al., 2011; Lunardi et al., 2014) mostly due to increased chemoresistance (Delitto et al., 2015). However, the role of CXCL10/CXCR3 signaling during early development of the disease has not been addressed.
Using the KC mouse model, we show that CXCL10 is produced by cells of precancerous ADM and PanIN1 lesions (Figure 1A,B). CXCL10 is also known as interferon gamma-inducible protein-10 (IP-10) (Luster et al., 1985; Ohmori and Hamilton, 1993), suggesting IFNγ as a potential factor sufficient to drive CXCL10 expression. Indeed, SM3 PanIN cells showed upregulation of CXCL10 expression in response to IFNγ (Figure 1C–E). This was mediated through JAK-STAT1 signaling (Figure 1F), which has been previously described to be upregulated and activated via IFNγ in acinar cells (Gallmeier et al., 2005). Many cell types, including T cells and NK cells in the pancreatic microenvironment, have been shown to secrete IFNγ (Brauner et al., 2010; Castro et al., 2018; Corthay et al., 2005; Fogar et al., 2011), and analysis for CD4+, CD8+, and NKG2D+ cells indicated T cells and NK cells in the vicinity of lesions as a potential source for IFNγ in our model (Figure 1G).
Our in vitro data implicate peritoneal or bone-marrow-derived inflammatory macrophages as major recipients for CXCL10, where CXCL10 can act as a chemoattractant (Figure 2B, Figure 2—figure supplement 1A). This is important because influx of macrophages is a major driver of pancreatic inflammation (Gea-Sorlí and Closa, 2009). This correlates with in vivo data showing that in the KC model CXCR3 is expressed by pancreatic inflammatory macrophages, but not by Ym1+ M2 macrophages (Figure 2D). In patient tissue, we found that approximately 70% of CXCR3 expressing cells are inflammatory (CD68+;iNOS+) macrophages (Figure 2E,F).
Unlike other factors contributing to macrophage chemoattraction into the pancreas, such as sICAM1 (Liou et al., 2015), the expression of CXCL10 from acinar cells is not driven by the activation of oncogenic KRas (Figure 1—figure supplement 1D). This finding is supported by other studies that have shown CXCL10 upregulation in chronic pancreatitis patients, which occurs in the absence of oncogenic mutations (Singh et al., 2007).
The importance of CXCL10/CXCR3 signaling in inflammatory macrophages is underscored by the fact that blocking this interaction causes them to lose their identity, as evident by upregulation of markers for alternatively activated macrophages (Figure 3A–C). Even though neutralizing CXCR3 did not simultaneously decrease iNOS expression while increasing M2 markers (Figure 3A, Figure 3—figure supplement 1B), there was a marked drop in Cd38, a distinct marker for M1 polarized macrophages (Jablonski et al., 2015), indicating a loss in M1 identity (Figure 3B). Mechanistically, this shift in polarization upon CXCR3 neutralization may be explained by upregulation of the M2 transcription factor Irf4 and downregulation of the M1 transcription factor Irf5 (Figure 3—figure supplement 1D). Interestingly, the resulting alternatively activated macrophage population is characterized by expression of markers (Ym1, Fizz1, Arg1) that previously have been attributed to a subset of pancreatic macrophages (usually described as Ym1+ macrophages) that drive fibrinogenesis and tumor progression (Liou et al., 2017).
In support of this explant data, neutralizing CXCR3 in vivo also led to a decrease in inflammatory (F4/80+;pY701-STAT1+) macrophages (Figure 3C,D) and an increase in Ym1+ macrophages (Figure 4A,B). This accompanied increased fibrosis at the pancreatic lesion areas and more abnormal regions in the pancreas overall (Figure 4C, D, and F). Our data is in line with a recent study in which it was reported that Cxcr3−/− mice show an abundance of M2 macrophages in breast tumors (Oghumu et al., 2014). Specifically, similar to what we observed after CXCR3 neutralization, Cxcr3−/− macrophages from these mice had reduced ability to upregulate iNOS and showed a predisposition for M2 polarization (Oghumu et al., 2014). Since there are other ligands described that can engage the receptor CXCR3 (Van Raemdonck et al., 2015), we also neutralized CXCL10 in KC mice and found similar trends of increased presence of Ym1+ macrophages and increased fibrosis (Figure 4—figure supplement 1). In a reverse experiment, enhanced expression of CXCL10 in PanIN cells significantly decreased overall abnormal lesion areas, fibrosis, and presence of M2 macrophages, but increased the presence of M1 macrophages (Figure 5).
CXCL10 is known to chemoattract cell types other than macrophages, mainly T cells, and approximately 15% of pancreatic T cells in KC mice are positive for CXCR3 (Figure 2C). However, our in vivo study in which we overexpressed CXCL10 in PanIN cells and observed reverse effects to CXCR3 or CXCL10 neutralization was performed in nude mice which lack T cells (Figure 4—figure supplement 2). In addition, depletion of CD4+ and CD8+ T cells in immunocompetent KC mice had no impact on total abnormal tissue area or M1 and M2 macrophage abundance after CXCR3 neutralization (Figure 4—figure supplement 3A–E). Taken together these data suggest no significant role for T cells in CXCL10/CXCR3-driven development of pancreatic cancer.
During pancreatic cancer development, inflammatory macrophages are predominantly located at ADM lesions, driving the transdifferentiation process (Liou et al., 2015; Liou et al., 2013). Our data now indicate that cells undergoing ADM (and to some extent low-grade PanIN) express CXCL10 to chemoattract inflammatory macrophages and that CXCL10/CXCR3 signaling is crucial for the sustenance of their inflammatory identity (Figure 5F). More progressed PanIN1 lesions have been shown to express the anti-inflammatory cytokine interleukin-13 (IL-13), which causes a polarization shift from M1 toward an alternatively activated, Ym1+ phenotype (Liou et al., 2017). Given the importance of CXCL10/CXCR3 interaction in the maintenance of the inflammatory identity of macrophages in our study, the reduced CXCL10 expression by PanIN as compared to ADM lesions (Figure 1A,B) may leave macrophages more susceptible to M2 polarization by cytokines like IL-13.
Considering our results, use of agonists for the receptor CXCR3 at stages of low-grade lesions may be useful to modulate macrophage polarization in the microenvironment such that a predominantly inflammatory population (M1) can be sustained. However, it needs to be noted that in pancreatic cancer expression of both CXCL10 and CXCR3 in tumor tissue have been correlated with metastasis and poor prognosis (Cannon et al., 2020; Hirth et al., 2020; Liu et al., 2011; Lunardi et al., 2014; Romero et al., 2020). Therefore, it is unclear if treatment of pancreatic tumors with a CXCR3 agonist will result in a polarization switch of tumor-associated macrophages that renders the lesion microenvironment less supportive for tumors, and increases efficiency of chemotherapy, or if it has a tumor-promoting effect. This will be addressed in future studies.
Materials and methods
Cells, antibodies, and reagents
Request a detailed protocolSM3 primary duct-like cells were isolated from pancreata of 6 week old KC mice as previously described (Agbunag et al., 2006; Liou et al., 2017). The genotype of these cells has been verified, and cells are routinely tested for mycoplasma infection. SM3 cells were maintained in DMEM/F12 media (Sigma-Aldrich, St. Louis, MO) containing 5% Nu Serum IV culture supplement (Corning, Corning, NY), 25 μg/ml bovine pituitary extract (Gibco/Thermo Scientific, Waltham, MA), 20 ng/ml EGF, 0.1 mg/ml soybean trypsin inhibitor type I (AMRESCO, Solon, OH), 5 mg/ml d-glucose (Sigma-Aldrich), 1.22 mg/ml nicotinamide (Sigma-Aldrich), 5 nM triiodo-l-thyronine (Sigma-Aldrich), 1 μM dexamethasone (Sigma-Aldrich), 100 ng/ml cholera toxin (Sigma-Aldrich), 5 ml/l insulin-transferrin-selenium (Corning), and 100 U/ml penicillin/streptomycin (Gibco/Thermo Scientific). Collagenase I was from MilliporeSigma (St. Louis, MO). Rat tail collagen I was from BD Biosciences (San Diego, CA). All antibodies used for western blotting, immunohistochemistry and immunofluorescence were from the following sources: Peprotech (Rocky Hill, NJ), Abcam (Cambridge, MA), Cell Signaling Technologies (Danvers, MA), Dako (Santa Clara, CA), STEMCELL Technologies (Vancouver, Canada), Biorad (Hercules, CA), and LifeSpan BioSciences (Seattle, WA), BD Biosciences (Franklin Lakes, NJ), BioLegend (San Diego, CA), GeneTex (Irvine, CA), Miltenyi Biotec (Auburn, CA) and are described in detail in Key Resources Table. All neutralizing antibodies that have been used in animal studies are described in detail in Materials and methods section for these experiments as well as in Key Resources Table. Secondary HRP-linked anti-mouse or anti-rabbit antibodies were from Jackson ImmunoResearch (West Grove, PA). DAPI and LPS were from Sigma-Aldrich. IL4, IFNγ, and CXCL10 were from PeproTech (Rocky Hill, NJ).
Isolation of primary macrophages (peritoneal or bone marrow derived) and polarization
Request a detailed protocolPeritoneal macrophages: Non-transgenic mice were injected with 2–3 ml of 5% aged thioglycollate. Five days later, mice were euthanized and peritoneal macrophages were collected immediately. A small incision was made in the peritoneum and RPMI-1640 media with 10% FBS was flushed through the peritoneal cavity using a sterile transfer pipette. The peritoneal wash was centrifuged, washed with fresh RPMI1640 + 10% FBS media, and cells were plated onto 10 cm dishes. After 2–3 hr, once macrophages completely adhered to the plates, cells were washed two to three times to remove non-adherent cells and fresh media was added. Cells were allowed to acclimatize overnight before harvesting and experimental manipulation. Bone marrow-derived macrophages (BMDM): Femurs of non-transgenic mice were cut proximal to the joints on both ends and washed twice in cold, sterile PBS (without calcium and magnesium) to remove excess blood and muscle. Using forceps, the bones were held over a sterile 50 ml tube, and the bone cavity was flushed twice with 5 ml cold PBS using a 25G needle and syringe. The bones were discarded, and the bone marrow was pelleted (10 min, 500 × g, RT), then gently resuspended to single cells in 10 ml warm macrophage complete medium (DMEM-F12 with l-glutamine + 10% FBS + 100 U/ml penicillin + 100 µg/ml streptomycin + 100 U/ml of M-CSF) (Zhang et al., 2008). After counting, 4 × 105 cells were added to 10 cm dishes in 10 ml macrophage complete medium. On days 3 and 5, 5 ml of medium was replaced with fresh complete medium. On day 7, the cells were polarized. Polarization of peritoneal or bone marrow-derived macrophages to M1 or M2 phenotypes was obtained by stimulation with LPS + IFNγ (M1 polarization) and/or with IL-4 (M2 polarization) for 24 hr as previously described (Liou et al., 2017).
Isolation of pancreas tissue resident macrophages
Request a detailed protocolPancreata from 8 week old, non-transgenic mice were harvested by washing the pancreas in HBSS, mincing the tissue, centrifuging (931 × g, 4°C, 2 min), and then dissociating the tissue in collagenase (2 mg/ml, 5 ml, 37°C, 20 min, 220 rpm). To stop dissociation, HBSS + 5% FBS was added to the pancreas–collagenase mixture and washed two additional times with HBSS + 5% FBS (931 × g, 4°C, 2 min). Pancreas cells were then filtered through 500 μm and 105 μm meshes prior to adding the cell suspension to HBSS + 30% FBS and centrifuging (233 × g, 4°C, 2 min) to obtain a cell pellet. Cells were then labeled with F4/80 magnetic beads (Miltenyi Biotec, Auburn, CA) to isolate pancreas-resident macrophages. Cells were filtered through a 40 μm mesh to obtain a single-cell suspension, and then cells were incubated with F4/80 magnetic beads in MACS buffer (15 min, 4°C, as per the manufacturer’s instructions) (MACS buffer: PBS, 0.5% BSA, 2 mM EDTA). Cells were washed with MACS buffer (300 × g, 4°C, 10 min), filtered through 40 μm mesh, and then applied to pre-washed LS Columns (Miltenyi Biotec, Auburn, CA). Per column instructions, magnetically labeled cells were acquired and plated in DMEM-F12 + 10% FBS + 1% l-glutamine + penicillin/streptomycin + 0.1 μg/ml macrophage colony stimulating factor (Peprotech, Rocky Hill, NJ).
Genetic animal model and treatments
Request a detailed protocolPtf1a/p48cre/+ and LSL-KrasG12D/+ mouse strains and genotyping of mice have been described previously (Liou et al., 2015). Seven to 8 week old Ptf1a/p48cre;LSL-KrasG12D (KC) or non-transgenic (ntg) mice with the same background were injected intraperitoneally (IP) with a CXCR3 neutralizing antibody (CXCR3 NAB; BE0249) (Bio X Cell, West Lebanon, NH) or an isotype control IgG; BE0091 (Bio X Cell) at 200 µg/mouse for 9 weeks. Males and females were randomly allocated to different groups since there are no sex-based differences observed in this model. All animal experiments were conducted under IACUC-approved protocols (A50214-14-R17, A30615-15-R18) and were run in accordance with institutional guidance and regulation.
For T-cell depletion, 6 week old Ptf1a/p48cre;LSL-KrasG12D mice were intraperitoneally injected with both anti-mouse CD4 (BP0003, Bio X Cell) and anti-mouse CD8α (BP0061, Bio X Cell) antibodies, or their IgG2b isotype control (BP0090, Bio X Cell). Two hundred micrograms of each antibody was injected per mouse for five consecutive days. After T-cell depletion, mice were segregated into different groups and injected with CXCR3 NAB (BE0249, Bio X Cell) or IgG isotype control antibodies (BE0091, Bio X Cell).
Human pancreatic tissue samples
Request a detailed protocolPatient tissues were obtained from archival materials in accordance with institutional guidelines and prior institutional review board (IRB) approval.
Immunofluorescence on cells and tissue
Request a detailed protocolFor paraffin-embedded tissue, slide sections were deparaffinized, and antigen retrieval was performed with 10 mM sodium citrate buffer (pH 6.0) for 25 min at 100°C. Then, they were treated with 3% H2O2 for 15 min at room temperature, rinsed with PBS, and blocked with serum-free Protein Block (DAKO, Santa Clara, CA) for 1 hr at room temperature. Slides were incubated with primary antibodies diluted in 150 µl per slide of Antibody Diluent (Dako) overnight. Slides were washed three times with 0.5% Tween20 in PBS and incubated with appropriate Alexa-Fluor labeled secondary antibodies (Invitrogen, Carlsbad, CA) at 1:500 and DAPI at 125 µg/ml for 1 hr. Slides were washed three times with 0.5% Tween20 in PBS followed by cover-slipping with PermaFluor mounting media (Thermo Fisher Scientific, Waltham, MA). For imaging, whole slides were scanned using the Aperio fluorescence scanner (Leica, Buffalo Grove, IL) or the Pannoramic 250 Flash III (3DHISTECH).
For cellular staining, cells grown in ibidi chamber slides were washed with cold PBS and fixed with 4% paraformaldehyde (PFA) for 15 min at 37°C. Fixed cells were washed with PBS and permeabilized with 0.1% Triton X-100 for 10 min at room temperature. Then, cells were blocked with 3% BSA + 0.05% Tween in PBS for 30 min at room temperature. Incubation with primary antibodies was done overnight at 4°C. Cells were washed with PBS and incubated with secondary antibodies at 1:500 and DAPI at 125 µg/ml for 2 hr. Cells were washed, and mounting media was added to the wells, followed by imaging on a fluorescence microscope (Carl Zeiss, Thornwood, NY).
Flow cytometry
Request a detailed protocolMurine pancreata were resected, and tissue was dissociated using the mouse tumor dissociation kit (Miltenyi, Bergisch Gladbach, Germany) per the manufacturer’s protocol. Dissociated tissue was sequentially passed through 500 μm (Repligen, Boston, MA), 105 μm (Repligen), and 40 μm (Thermo Fisher Scientific) filters to acquire a single-cell suspension before magnetic isolation of CD45+ cells. Cells were incubated with CD45+ MicroBeads (Miltenyi) followed by magnetic separation using LS Columns (Miltenyi) on a QuadroMACS magnet (Miltenyi) per the manufacturer’s protocol. CD45+ cells were then labeled with LIVE/DEAD Fixable Violet Dead Cell Stain Kit (Thermo Fisher Scientific) before subsequent fixation and staining using the PerFix-nc kit (Beckman Coulter, Brea, CA) as per the manufacturer’s protocol. Antibodies are described in detail in Key Resources Table. The Attune NxT (Thermo Fisher Scientific) was used for multicolor flow cytometric detection, and analysis was done using FlowJo (BD Biosciences).
In situ hybridization
Request a detailed protocolThe procedure for ISH has been described in detail previously (Bastea et al., 2019). Briefly, ISH was performed using RNAscope Assay 2.5 HD Reagent Kit–Brown or RNAscope Multiplex Fluorescent Reagent Kit v2 (Advanced Cell Diagnostics, Hayward, CA). Formalin-fixed, paraffin-embedded (FFPE) sections (5 μm) were baked at 60°C for 1 hr, deparaffinized in xylene for 15 min, dehydrated in 100% ethanol, and dried at room temperature overnight in a desiccator. Next day, slides were treated with hydrogen peroxide for 10 min, followed by target retrieval for 8 min. Protease treatment was done for 15 min at 40°C in a hybridization oven. Next, slides were incubated with the appropriate probe for 2 hr. RNAscope target probes used were Cxcl10 (Mm-Cxcl10 408921), Cxcr3 (Mm-Cxcr3 4025110), CXCR3 (Hs-CXCR3 539251), and Ifng (Mm-Ifng 311391). After probe hybridization, amplification steps were followed according to the manufacturer’s protocol, except Amp 5 step, which was modified to 1 hr incubation in the DAB procedure (RNAscope Assay 2.5 HD Reagent Kit–Brown). Slides were counterstained with hematoxylin, dehydrated, and mounted.
When followed by IF, slides were treated with boiling 10 mM citrate buffer (pH 6.0) for 10 min to repeat antigen retrieval and the regular IF protocol was followed. Primary antibodies were used at concentrations listed in Key Resources Table.
For the fluorescent detection (RNAscope Multiplex Fluorescent Reagent Kit v2), following AMP1, 2 and 3 steps according to the manufacturer’s protocol, probes were labeled using the Opal 690 fluorophore reagent pack (Perkin Elmer, Waltham, MA). Next, regular IF procedure was performed for co-staining. Samples were counterstained with DAPI and mounted. For imaging, whole slides were scanned at 40× magnification using the Aperio fluorescence scanner (Leica).
Isolation of primary acinar cells
Request a detailed protocolPancreas was removed quickly following euthanasia and placed in cold HBSS media. After washing in cold HBSS media twice, pancreas was minced into small 1–5 mm pieces and digested with collagenase I in a 37°C shaker for 20 min. After 20 min, the digestion was stopped by adding an equal volume of cold HBSS media with 5% FBS. The digested pieces were passed through a 500 μm mesh. More 5% FBS media was passed through the mesh with digested pieces to further strain them. The filtrate was then passed through a 105 μm mesh. The filtrate from this step was added dropwise to a tube containing HBSS media with 30% FBS. The cell suspension was centrifuged at 205 × g for 2 min at 4°C. Acinar cells were then resuspended in complete Waymouth’s media (1% FBS, 0.1 mg/ml soybean trypsin inhibitor, 1 μg/ml dexamethasone).
3D collagen explant culture of pancreatic acinar cells
Request a detailed protocolCell culture plates were coated with collagen I in Waymouth media without supplements. Freshly isolated primary pancreatic acinar cells were added as a mixture with collagen I/Waymouth media on the top of this layer (3D on-top method). Furthermore, Waymouth complete media was added on top of the cell/gel mixture and replaced the following day and then every other day. When inhibitors, peptides, or proteins were added, the compound of interest was added to both, the cell/gel mixture and the media on top. To express proteins using adenovirus, acinar cells were infected with adenovirus of interest and incubated for 3–5 hr before embedding in the collagen I/Waymouth media mixture. At day 6 or 7 (dependent on time course of duct formation), numbers of ducts were counted under a microscope, and photos were taken to document structures.
Adenoviral infection of acinar cells
Request a detailed protocolFor adenoviral infection, Adeno-cre-GFP or Adeno-null-GFP viruses (Vector Biolabs, Malvern, PA) were added to the resuspended acinar cells in Waymouth’s complete media in a non-adherent dish. Infection was carried out for 3 hr with gentle swirling every 15 min during the first hour. After the 3 hr infection, cell suspension was mixed with Rat tail collagen-Type I (BD Biosciences, San Jose, CA) and plated.
Lentiviral infection of PanIN lesion cells, organoid formation, and orthotopic implantation in athymic nude mice
Request a detailed protocolSM3 PanIN cells were infected using control lenti-eGFP (LPP-EGFP-Lv105-025-C) or lenti-cxcl10 (LPP-Mm03214-Lv105-100) lentiviral particles (Genecopoeia, Rockville, MD) with 5 μg/ml of polybrene (Santa Cruz Biotechnology, Dallas, TX) at 70–80% confluency. Infected cells were incubated at 4°C for 2 hr and then at 37°C overnight. On the following day, media containing lentiviral particles and polybrene was removed and replaced with fresh media to allow the cells to recover. Next day, selection media containing 2.5 µg/ml puromycin was added to the cells and was replaced every 2–3 days. After 11 days of selection, cells were harvested and embedded in Matrigel (Corning, Corning, NY) with regular media without puromycin. After 2 days, cells formed duct-like PanIN organoid structures. Using a non-enzymatic cell dissociation reagent (Corning), PanIN organoids were harvested and mixed with activated primary pancreatic stellate cells at a ratio of 1:4. Cell mixture was resuspended in phenol-red free Matrigel, and 50 µl (25,000 organoid cells + 100,000 activated primary stellate cells) was injected directly into the pancreas of athymic nude (Foxn1nu) mice (Jackson laboratory, Bar Harbor, ME). Wound was closed using 4–0 Vicryl sutures and 7 mm wound clips. After 1 week of recovery, wound clips were removed, and ultrasound imaging was performed to ensure implantation of cells.
RNA extraction and quantitative PCR
Request a detailed protocolCells were washed with cold PBS, and RNA extraction was done using the RNeasy PLUS Mini Kit (Qiagen, Germantown, MD). cDNA was prepared using the High Capacity cDNA RT Kit (Applied Biosystems, Foster City, CA). TaqMan Fast Mix 2x (Applied Biosystems) was used to prepare qPCR along with the primer/probe sets described in Key Resources Table. Reactions were run on the QuantStudio 7 Flex Real-Time PCR System (Applied Biosystems). All CT values were normalized to Gapdh, and the ΔΔCT method was used to calculate fold changes.
Protein isolation and western blot
Request a detailed protocolCells were washed with cold PBS (140 mM NaCl, 2.7 mM KCl, 8 mM Na2HPO4, and 1.5 mM KH2PO4 [pH 7.2]) twice and lysed using RIPA buffer with a protease inhibitor cocktail (Sigma-Aldrich). Lysates were incubated on ice for 30 min followed by centrifugation at 13,000 rpm for 15 min at 4°C on a table top centrifuge. Protein supernatant was separated from cell debris, and concentration was measured using the BioRad Protein Assay (BioRad, Hercules, CA). Samples were run on SDS-PAGE gels and transferred to nitrocellulose membranes. Proteins of interest were detected using the appropriate primary antibodies at indicated concentrations (Key Resources Table) and horseradish peroxidase (HRP)-conjugated secondary antibodies.
Transwell chemoattraction assays
Request a detailed protocolPeritoneal or bone marrow-derived macrophages were seeded in serum-free RPMI media on 5 μm transwell permeable inserts (Corning, Corning, NY). Six hundred microliters of control media or media containing 500 ng/ml of CXCL10 was placed in the bottom wells. Experiments were conducted in triplicates for each condition. Cells were incubated for the indicated time. Inserts were then carefully removed, washed with PBS, and fixed in 4% PFA for 15 min at 37°C. After gently washing the inserts twice, cells were permeabilized with 0.1% Triton X-100 for 10 min at room temperature and stained with DAPI for counting.
MTT proliferation assay
Request a detailed protocolProliferation of pancreas-resident macrophages was analyzed via MTT assay (Sigma, St. Louis, MO), where wells were incubated with MTT labeling reagent (4 hr, 37°C, 5% CO2). Following MTT labeling, cells were incubated with solubilization solution (10% SDS in 0.01 M HCl) overnight (37°C, 5% CO2), and absorbance was read the following morning at 500 nm (Synergy HT plate reader, BioTek, Winooski, VT).
Cytokine array
Request a detailed protocolThe cytokines secreted by SM3 cells were determined using the Proteome Profiler Mouse Cytokine Array Kit (R and D Systems) according to the manufacturer’s instructions.
Quantification and statistical analysis
Request a detailed protocolAll cell biological and biochemical experiments have been performed at least three times. For animal experiments, if not stated otherwise in the figure legends, pancreatic samples from n = 3 mice have been used for quantification analyses. IHC data was quantified by manual counting of positive cells or by using the Aperio Positive Pixel Count Algorithm. Data are presented as mean ± standard deviation (SD). If not stated otherwise in the figure legends, p-values were acquired with the unpaired student’s t-test with Welch’s correction using Graph Pad software (GraphPad Inc, La Jolla, CA). For all experiments in which pancreatic areas were compared, we transformed the data via arcsin transformation (ASIN(SQRT(proportion of abnormal tissue area to total tissue area))*180/PI) before a t-test was performed. p<0.05 was considered statistically significant.
Appendix 1
Data availability
All data generated or analysed during this study are included in the manuscript and supporting files. Source data have been included for all Figures and Figure supplements.
References
-
Distinct phenotype and function of NK cells in the pancreas of nonobese diabetic miceThe Journal of Immunology 184:2272–2280.https://doi.org/10.4049/jimmunol.0804358
-
Interferon-Gamma at the crossroads of tumor immune surveillance or evasionFrontiers in Immunology 9:847.https://doi.org/10.3389/fimmu.2018.00847
-
B7-H3: a costimulatory molecule for T cell activation and IFN-gamma productionNature Immunology 2:269–274.https://doi.org/10.1038/85339
-
The chemokine receptors CXCR4 and CXCR3 in CancerCurrent Oncology Reports 11:125–131.https://doi.org/10.1007/s11912-009-0019-1
-
JAK and STAT proteins are expressed and activated by IFN-γ in rat pancreatic acinar cellsJournal of Cellular Physiology 203:209–216.https://doi.org/10.1002/jcp.20216
-
CXCR3 ligands: redundant, collaborative and antagonistic functionsImmunology & Cell Biology 89:207–215.https://doi.org/10.1038/icb.2010.158
-
Interferon-Regulatory factors determine macrophage phenotype polarizationMediators of Inflammation 2013:1–8.https://doi.org/10.1155/2013/731023
-
Targeting chemokines and chemokine receptors in melanoma and other cancersFrontiers in Immunology 9:2480.https://doi.org/10.3389/fimmu.2018.02480
-
In vivo depletion of T lymphocytesCurrent Protocols in Immunology 113:1–9.https://doi.org/10.1002/0471142735.im0401s113
-
Impact of chemokine receptor CXCR3 on tumor-infiltrating lymphocyte recruitment associated with favorable prognosis in advanced gastric CancerInternational Journal of Clinical and Experimental Pathology 8:14725–14732.
-
Macrophage-secreted cytokines drive pancreatic acinar-to-ductal metaplasia through NF-κB and MMPsJournal of Cell Biology 202:563–577.https://doi.org/10.1083/jcb.201301001
-
The emerging role of CXCL10 in Cancer (Review)Oncology Letters 2:583–589.https://doi.org/10.3892/ol.2011.300
-
Biochemical characterization of a gamma interferon-inducible cytokine (IP-10)Journal of Experimental Medicine 166:1084–1097.https://doi.org/10.1084/jem.166.4.1084
-
Targeting the tumor microenvironment in pancreatic ductal adenocarcinomaExpert Review of Anticancer Therapy 19:473–482.https://doi.org/10.1080/14737140.2019.1622417
-
CXCL10 expression status is prognostic in patients with advanced thoracic esophageal squamous cell carcinomaAnnals of Surgical Oncology 23:936–942.https://doi.org/10.1245/s10434-015-4909-1
-
Cellular determinants and therapeutic implications of inflammation in pancreatic CancerPharmacology & Therapeutics 201:202–213.https://doi.org/10.1016/j.pharmthera.2019.05.012
-
Acinar cell plasticity and development of pancreatic ductal adenocarcinomaNature Reviews Gastroenterology & Hepatology 14:296–304.https://doi.org/10.1038/nrgastro.2017.12
-
CXCR3 ligands in disease and therapyCytokine & Growth Factor Reviews 26:311–327.https://doi.org/10.1016/j.cytogfr.2014.11.009
-
The isolation and characterization of murine macrophagesCurr Protoc Immunol Chapter 14:11.https://doi.org/10.1002/0471142735.im1401s83
-
CXCR3-dependent accumulation and activation of perivascular macrophages is necessary for homeostatic arterial remodeling to hemodynamic stressesJournal of Experimental Medicine 207:1951–1966.https://doi.org/10.1084/jem.20100098
Article and author information
Author details
Funding
National Cancer Institute (CA229560)
- Peter Storz
National Cancer Institute (CA200572)
- Peter Storz
National Cancer Institute (P50CA102701)
- Peter Storz
The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.
Acknowledgements
We thank Dr. Yan Asmann (Associate Professor of BioMedical Informatics, Mayo Clinic) for advice on the statistical analyses. We also thank our colleagues in the Storz laboratory and Department for Cancer Biology for helpful discussions. We also thank the Champions For Hope (Funk-Zitiello Foundation) and Gary Chartrand Foundation for their continuous support.
Ethics
Animal experimentation: This study was performed in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. All animal experiments were conducted under IACUC approved protocols (A50214-14-R17, A30615-15-R18) and were run in accordance with institutional guidance and regulation.
Copyright
© 2021, Pandey et al.
This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.
Metrics
-
- 4,326
- views
-
- 543
- downloads
-
- 44
- citations
Views, downloads and citations are aggregated across all versions of this paper published by eLife.
Download links
Downloads (link to download the article as PDF)
Open citations (links to open the citations from this article in various online reference manager services)
Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)
Further reading
-
- Cancer Biology
- Computational and Systems Biology
Effects from aging in single cells are heterogenous, whereas at the organ- and tissue-levels aging phenotypes tend to appear as stereotypical changes. The mammary epithelium is a bilayer of two major phenotypically and functionally distinct cell lineages: luminal epithelial and myoepithelial cells. Mammary luminal epithelia exhibit substantial stereotypical changes with age that merit attention because these cells are the putative cells-of-origin for breast cancers. We hypothesize that effects from aging that impinge upon maintenance of lineage fidelity increase susceptibility to cancer initiation. We generated and analyzed transcriptomes from primary luminal epithelial and myoepithelial cells from younger <30 (y)ears old and older >55y women. In addition to age-dependent directional changes in gene expression, we observed increased transcriptional variance with age that contributed to genome-wide loss of lineage fidelity. Age-dependent variant responses were common to both lineages, whereas directional changes were almost exclusively detected in luminal epithelia and involved altered regulation of chromatin and genome organizers such as SATB1. Epithelial expression of gap junction protein GJB6 increased with age, and modulation of GJB6 expression in heterochronous co-cultures revealed that it provided a communication conduit from myoepithelial cells that drove directional change in luminal cells. Age-dependent luminal transcriptomes comprised a prominent signal that could be detected in bulk tissue during aging and transition into cancers. A machine learning classifier based on luminal-specific aging distinguished normal from cancer tissue and was highly predictive of breast cancer subtype. We speculate that luminal epithelia are the ultimate site of integration of the variant responses to aging in their surrounding tissue, and that their emergent phenotype both endows cells with the ability to become cancer-cells-of-origin and represents a biosensor that presages cancer susceptibility.
-
- Cancer Biology
Metastasis is the leading cause of cancer-related mortality. Paneth cells provide stem cell niche factors in homeostatic conditions, but the underlying mechanisms of cancer stem cell niche development are unclear. Here, we report that Dickkopf-2 (DKK2) is essential for the generation of cancer cells with Paneth cell properties during colon cancer metastasis. Splenic injection of Dkk2 knockout (KO) cancer organoids into C57BL/6 mice resulted in a significant reduction of liver metastases. Transcriptome analysis showed reduction of Paneth cell markers such as lysozymes in KO organoids. Single-cell RNA sequencing analyses of murine metastasized colon cancer cells and patient samples identified the presence of lysozyme positive cells with Paneth cell properties including enhanced glycolysis. Further analyses of transcriptome and chromatin accessibility suggested hepatocyte nuclear factor 4 alpha (HNF4A) as a downstream target of DKK2. Chromatin immunoprecipitation followed by sequencing analysis revealed that HNF4A binds to the promoter region of Sox9, a well-known transcription factor for Paneth cell differentiation. In the liver metastatic foci, DKK2 knockout rescued HNF4A protein levels followed by reduction of lysozyme positive cancer cells. Taken together, DKK2-mediated reduction of HNF4A protein promotes the generation of lysozyme positive cancer cells with Paneth cell properties in the metastasized colon cancers.