SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

Natriuretic peptide receptor A (NPRA), the signaling receptor for the cardiac hormone, atrial natriuretic peptide (ANP), is expressed abundantly in inflamed/injured tissues and tumors. NPRA deficiency substantially decreases tissue inflammation and inhibits tumor growth. However, the precise mechanism of NPRA function and whether it links inflammation and tumorigenesis remains unknown. Since both injury repair and tumor growth require stem cell recruitment and angiogenesis, we examined the role of NPRA signaling in tumor angiogenesis as a model of tissue injury repair in this study. In in vitro cultures, aortas from NPRA-KO mice show significantly lower angiogenic response compared to wild-type counterparts. The NPRA antagonist that decreases NPRA expression, inhibits lipopolysaccharide-induced angiogenesis. The reduction in angiogenesis correlates with decreased expression of vascular endothelial growth factor and chemokine (C-X-C motif) receptor 4 (CXCR4) implicating a cell recruitment defect. To test whether NPRA regulates migration of cells to tumors, mesenchymal stem cells (MSCs) were administered i.v., and the results showed that MSCs fail to migrate to the tumor microenvironment in NPRA-KO mice. However, coimplanting tumor cells with MSCs increases angiogenesis and tumorigenesis in NPRA-KO mice, in part by promoting expression of CXCR4 and its ligand, stromal cell-derived factor 1α. Taken together, these results demonstrate that NPRA signaling regulates stem cell recruitment and angiogenesis leading to tumor growth. Thus, NPRA signaling provides a key linkage between inflammation and tumorigenesis, and NPRA may be a target for drug development against cancers and tissue injury repair. STEM Cells2013;31:1321–1329


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

Atrial natriuretic peptide (ANP) is the C-terminal fragment of pro-ANP. Its receptor, natriuretic peptide receptor A (NPRA), is expressed on cells in inflamed/injured tissues and in tumors [1, 2]. NPRA signals through guanylyl cyclase by increasing cyclic guanosine 3′, 5′-monophosphate (cGMP) and activates cGMP-dependent protein kinase (PKG) [3, 4], which in turn, upregulates gene expressions affecting cell proliferation and inflammation. NPRA deficiency substantially decreases tissue inflammation and inhibits tumor growth [1]. However, the precise mechanism of NPRA function and whether it links inflammation and tumorigenesis remains unknown.

As in tissue injury repair, tumor growth requires stem or progenitor cell recruitment [5], their differentiation into other cell types including endothelial progenitor cells (EPCs) [6, 7], which differentiate into endothelial cells (ECs) and form new capillaries leading to tumor growth. The tumor vasculature is less organized and leakier than normal vasculature [8] and once the new blood vessels are formed, they aid in further tumor growth and metastasis to different regions of the body [9]. Crosstalk between tumor cells and surrounding stromal cells results in secretion of extracellular matrix proteins, growth factors, cytokines, and chemokines including the vascular endothelial growth factor (VEGF) for inducing blood vessel formation [10, 11] and stromal cell-derived factor 1 alpha (SDF1α) that produce oncogenic signals, and increase tumor survival, angiogenesis, invasion, and growth [12–15]. SDF1α acts as a chemoattractant for stem cells, and EPCs, which express (C-X-C motif) receptor 4 (CXCR4), the receptor for SDF1α [16, 17]. While the role of CXCR4-SDF1 in the cell recruitment per se has been reported, the upstream regulator of such interaction remains to be elucidated.

We have reported that NPRA is an early biomarker for human prostate cancer [2] and has the potential for clinical staging of the disease. Furthermore, we recently established NPRA as a biomarker for melanoma, colon, and pancreatic cancer (Supporting Information Fig. S1). Given the importance of NPRA as an anticancer drug target, we investigated the potential mechanism for the role of NPRA in cancer pathogenesis. Since, ANP have been implicated in local inflammation, we reasoned that NPRA signaling provides an excellent model to study link between inflammation and tumorigenesis. NPRA signaling may promote tumorigenesis by influencing recruitment of immune and progenitor cells, and thereby fostering angiogenesis in the tumor microenvironment (TME). To test this idea, in this study we examined the recruitment of stem cell progenitors to the microenvironment of tumors grown in NPRA-knock out (NPRA-KO) mice. In addition, we investigated intrinsic and induced angiogenesis after attenuating NPRA signaling. The results of our studies reported herein for the first time provide evidence that NPRA signaling plays a pivotal role in regulating both intrinsic and inflammation-induced angiogenesis required for tumor growth. Specifically, NPRA signaling modulates the inflammation in the TME by controlling recruitment of progenitor cells that are critical for tumor growth.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

Reagents

Growth factor-reduced Matrigel was purchased from BD Biosciences (San Jose, CA, http://www.bdbiosciences.com). Medium 200PRF, Dulbecco's modified Eagle's medium (DMEM), and low serum growth supplements were obtained from Invitrogen (Carlsbad, CA, http://www.invitrogen.com). Anantin was obtained from Bachem (Torrance, CA), lipopolysaccharide (LPS) from Bachem, optimum cutting temperature (OCT) and fetal bovine serum (FBS) from Fisher Scientific (Waltham, MA, http://www.fisherscientific.com). Polyclonal antibody to CXCR4 was purchased from R&D (Minneapolis, MN, http://www.rndsystems.com) and to NPRA, SDF1α, α-smooth muscle actin (SMA), S100A4, and von Willebrand factor (vWF) and monoclonal antibodies to CD31 (P2B1), F4/80 (A3-1) antibodies were purchased from Abcam (Cambridge, MA, http://www.abcam.com). Monoclonal antibody to vimentin (D2-H3) was purchased from Cell Signaling (Danvers, MA, http://www.cellsignal.com) and carcinoembryonic antigen (CEA) was purchased from Genemed (San Francisco, CA). Isotype-matched primary antibodies were purchased from Sigma (St. Louis, MO, http://www.sigmaaldrich.com). Labeled secondary antibodies were obtained from Invitrogen.

Mice

C57BL/6 wild-type (WT) mice were obtained from the National Cancer Institute. Green fluorescent protein (GFP)-transgenic mice were obtained from Jackson Labs. NPRA-KO mice were bred and genotyped in-house, as described [1, 2]. All mice were maintained in a pathogen-free environment with 12-hour day/night cycle.

Quantitation of VEGF and SDF1α

Supernatants from aortic ring cultures and tumor explants were tested for VEGF (Peprotech, Rocky Hill, NJ, http://www.peprotech.com) and SDF1α (Ray Biotech, Norcross, GA) by ELISA according to the manufacturers' instructions. Data were converted to ng/mL using a standard curve generated by linear regression and analyzed for statistical significance.

Mesenchymal Stem Cell Isolation and Culture

Bone marrow cells were flushed from the long bones of mice with DMEM supplemented with 10% FBS. The cells obtained were plated at a density of 1.5 × 106 per cm2 and incubated for 3–4 days. Mesenchymal stem cells (MSCs) were isolated based on adherence to plastic surfaces, while floating blood and hematopoietic stem cells were gently aspirated. The attached MSCs were grown to 80% confluence, phenotyped by flow cytometry using a mouse multipotent MSC multi-color flow cytometry kit (R&D, Minneapolis, MN, http://www.rndsystems.com), and subcultured until passage 10 for the experiments.

Mouse Aortic Ring Assay

Aortic ring assay was performed as per method of Bellacen and Lewis [18]. Mouse aortas were isolated, cleared of periadventitial fat, and cut into rings of 1–1.5 mm in length. Thawed Matrigel (100 μL) was added to the wells of a 96-well plate kept on ice. The aortic rings were then sequentially randomized into Matrigel-filled wells, and the Matrigel was allowed to gel in an incubator at 37°C. Rings were incubated in basal medium for 1 day to normalize the angiogenic potential of the aortic rings. From the second day, the rings were incubated with growth medium (DMEM + 10% FBS) with replenishment on alternate days for 7–10 days. In the experiments involving MSCs, 2–3 × 103 MSCs were added over the Matrigel on the second day. After 7 days, the sprouted microvessels were imaged under the microscope. ImageJ was used to transform the aortic ring images to increase the distinction of the capillary growth from the background. The area of capillary growth was measured with ImageJ and statistically compared between the groups.

Cell Isolation from Mouse Aorta

Aortic cells were obtained from WT and NPRA-KO mice. Mouse aortas were dissected clean of the surrounding matrix and divided into 8–10 mm segments. The segments were cut open and sliced longitudinally into three to four sections. Then, one to two sections per well were placed in a 96-well plate and incubated in 200 μL of DMEM + 10% FBS. Cells from the aortic sections migrated onto the plate and once the cells were well-established, the aortic sections were gently removed. When the cells were 80% confluence, they were harvested and subcultured. To assess the tube forming ability, the aortic cells were plated over Matrigel at a density of 2 × 103 per well. After 16 hours, calcein AM was used to stain the capillary tube networks, which were imaged under the microscope and tube areas were quantified with ImageJ and statistically compared by Student's t test.

Lewis Lung Carcinoma Mouse Lung Tumor Model

WT female C57BL/6 (National Cancer Institute) and C57BL/6 NPRA-KO (bred in-house) mice were used to establish lung tumors with Lewis lung carcinoma (LLC-1) cells as described previously [1]. On day 1, 5 × 105 LLC-1 cells (American Type Culture Collection) were subcutaneously (s.c.) injected into the flanks of mice with or without MSCs. The tumors were monitored for 3 weeks, after which they were excised and weighed. To analyze the recruitment of MSCs to the TME, GFP-expressing MSCs (1 × 106) in 100 μL were injected i.v. into the tail vein on day 10. Tumors from all experiments were harvested on day 21, embedded in OCT, and stored at −80°C for cryosectioning and immunohistochemical analysis. To measure intratumoral cytokines, the tumors were minced, incubated at 37°C for 24 hours, tumor explant supernatants (TES) were collected and analyzed by ELISA.

Immunohistochemistry

The tumors and the aortic rings were sectioned with a cryostat and the sections were fixed with 4% paraformaldehyde in phosphate-buffered saline (PBS) for 15 minutes at room temperature, followed by permeabilization in 0.01% Triton X-100. The sections were incubated with 10% host serum that the secondary antibody was raised in for 30 minutes at room temperature. Sections were immunostained for macrophages (F4/80), endothelium (Tie2), endothelial progenitors (CD31), pericytes (SMA), GFP (GFP-MSCs), and CXCR4. Specificity of each of these antibodies was verified using isotype-matched control antibodies (Supporting Information Fig. S2A, S2B). For dual immunostaining, sections were incubated with the first primary antibody in 1% bovine serum albumin (BSA) in PBS plus Tween 20 (PBST) overnight at 4°C, washed three times with PBS, and incubated with the corresponding secondary antibody in 1% BSA in PBST for 1 hour at room temperature in dark. After washing three times with PBS, sections were blocked for a second time with 10% serum from the species the secondary antibody was raised in for 30 minutes at room temperature. Sections were then incubated with the second primary antibody in 1% BSA in PBST overnight at 4°C, followed by the second secondary antibody (labeled with Texas red) in 1% BSA for 1 hour at room temperature. The sections were stained with the nuclear stain 4',6-diamidino-2-phenylindole (DAPI) and imaged with a fluorescence microscope.

Statistical Analysis

Each experiment was repeated at least twice. ImageJ analyses were performed for measurement of capillary tube areas and cell counting. All statistical analyses were done in Graphpad Prism. Data are presented as means ± SEMs. The treated and untreated groups were compared using Student's t test. With more than two groups, analyses were carried out with one-way ANOVA, followed by the Bonferroni post hoc test. P values less than 0.05 were considered statistically significant. The stem cell differentiation represented by dual stain overlap was assessed by Mander's coefficient (R) for colocalization with Mander's coefficient plugin of ImageJ (NIH, Bethesda, MD).

Approval Declaration

All animal studies were approved and conducted according to guidelines of the University of South Florida Institutional Animal Care and Use Committee.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

NPRA Deficiency Inhibits Aortic Sprouting

To investigate the role of NPRA signaling in regulating angiogenesis, an ex vivo aortic sprout assay was established in which aortic rings embedded in Matrigel are grown for 7 days. We examined the angiogenic potential of the C57BL/6 WT and NPRA-KO mice aortas with aortic ring assay. Measurement of capillary growth from aortic rings showed that NPRA-KO mice had significantly lower capillary growth than WT (Fig. 1A, 1B). Moreover, the aortic cell capillary networks derived from NPRA-KO mice were more sparsely formed than those from WT mice (Fig. 1C). Histological analysis of lung sections of WT and KO mice did not show any statistically significant difference in the capillaries, however, significantly more vWF positive cells were found in the lung sections of the WT mice compared to NPRA-KO mice suggesting a reduced potential for blood vessel formation in the latter (Supporting Information Fig. S3). These findings suggest that NPRA signaling is important for blood vessel formation.

thumbnail image

Figure 1. NPRA signaling regulates capillary growth. (A, B): Comparison of angiogenic potential of WT and NPRA-KO mice. Aortas extracted from WT and NPRA-KO mice (n = 6 mice per group) were cultured for 7 days and imaged at a magnification of ×40 (A). The capillary growth areas from the aortic rings were measured with ImageJ, expressed as means ± SEM and statistically analyzed (B). (C): Capillary network formation by aortic cells from WT and NPRA-KO mice was assessed. Aortic cells from WT and NPRA-KO mice were plated on Matrigel and, after a 16-hour incubation, the capillaries were stained with calcein AM and imaged. Tube areas were measured by ImageJ, expressed as means ± SEM (n = 4 mice per group; 2 fields per mouse) and statistically compared. *, p < 0.05; ****, p < 0.0001. Abbreviations: KO, knock out; WT, wild type.

Download figure to PowerPoint

Since inflammatory mediators in the TME promote tumor growth, we examined the influence of NPRA signaling in LPS-promoted angiogenesis [15] using an aortic ring assay. WT aortic rings incubated with LPS exhibited more capillary growth than control (Fig. 2A, 2B). When the rings were incubated with LPS in the presence of an NPRA antagonist, anantin, no increase of capillary growth was observed. Aortic rings from NPRA-KO mice incubated with a same concentration of LPS failed to show capillary growth increase. The failure in capillary growth was correlated with decreased VEGF expression in these cultures (Fig. 2C). A higher concentration of VEGF protein was detected in the supernatants of LPS-treated WT rings than the control groups.

thumbnail image

Figure 2. NPRA signaling modulates LPS-promoted angiogenesis. Aortas removed from WT and NPRA-KO mice were incubated with or without (control; C) 10 μg/mL LPS (L) in growth medium with or without 1 μM anantin (A) and imaged on day 7 at a magnification of ×40 (A). The capillary growth areas were measured with ImageJ software, and values were expressed as means ± SEM (n = 3 aortic rings per group). The groups were statistically compared by one-way ANOVA (B). Supernatants collected from aortic ring cultures after incubating 7 days were also examined for VEGF concentration. Values are expressed as means ± SEM (n = 4 mice per group). Samples were pooled from three rings per group (C). *, p < 0.05. Abbreviations: KO, knock out; VEGF, vascular endothelial growth factor; WT, wild type.

Download figure to PowerPoint

NPRA Signaling Regulates Migration of the EPCs

Angiogenesis involves migration of stem cells, their differentiation to CD31-positive EPCs, followed by endothelial proliferation, migration, and capillary tube formation [5, 19, 20]. CXCR4, which interacts with SDF-1 secreted by cancer associated fibroblasts (CAFs) in the TME, is expressed on tumor cells, migrating stem cells, and ECs throughout the process of new blood vessel formation [21]. To test whether NPRA signaling regulates migration of the EPCs, we examined aortic sections for the expression of NPRA, CXCR4, and CD31. LPS treatment increased the expression of NPRA- and CXCR4-positive cells (Fig. 3) in the aortic sections of WT mice. Moreover, most of the CXCR4-positive cells were also positive for CD31. In contrast, treatment with anantin, which reduced NPRA expression, also inhibited the upregulation of CXCR4 and CD31. The capillaries of NPRA-KO mice also showed statistically significant reduction in CXCR4 and CD31 expression, and LPS treatment failed to increase their expression. Collectively, these results suggest that NPRA signaling is required for LPS-induced expression of NPRA and CXCR4 and the development of capillary tubes from aorta in vitro.

thumbnail image

Figure 3. NPRA signaling modulates CXCR4 and CD31 expression. Aortic rings embedded in Matrigel were cultured for 7 days in growth medium supplemented with the test reagents (as in Fig. 2) and then embedded in OCT, frozen, and cryosectioned. Sections were stained for NPRA (red), CXCR4 (green), or CD31 (red) and visualized using a fluorescence microscope at a magnification of ×400. Bar graphs show the number of NPRA, CXCR4, CD31 or CXCR4 and CD31 positive cells (n = 5 fields per group; one field per section per mouse) that were quantified with ImageJ and statistically compared by one-way ANOVA. The groups compared were C versus L; L versus L+A; WT (C) versus KO (C); KO (C) versus KO (L). *, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001. Scale bar = 20 μm. Abbreviations: CXCR4, chemokine (C-X-C motif) receptor 4; NPRA, natriuretic peptide receptor A; KO, knock out; WT, wild type.

Download figure to PowerPoint

NPRA-KO Mice Exhibit Impaired Stem Cell Recruitment to the TME

Tumor angiogenesis is essential for tumor growth because new blood vessels are required for the tumors to grow beyond 1 mm3. Cells that are recruited to the tumor site through angiogenesis, in turn, modulate angiogenesis. MSCs in the circulation migrate to tumors [22], where they differentiate to repair the damaged tissue. After homing to the tumor site, MSCs can differentiate to CAFs [23], ECs [24], and endothelial progenitors [24]. In addition, MSCs secrete growth factors and cytokines, which stimulate cell recruitment and angiogenesis that ultimately promotes tumor growth. Since NPRA-KO mice failed to induce inflammation [25–28] and propagate implanted tumors [1, 2], we reasoned that an inability to recruit bone marrow-derived stem cells may impair tumor angiogenesis and tumorigenesis. To investigate this we compared the recruitment of intravenously injected MSCs to the preimplanted subcutaneous tumors in the flanks of WT and NPRA-KO mice. LLC-1 cells were selected for tumor implantation in this study because they are highly tumorigenic in C57BL/6 mice and these cells express NPRA abundantly [1]. LLC-1 cells were injected subcutaneously into the flanks of WT and NPRA-KO mice and 10 days later, GFP-expressing MSCs were injected via the tail vein. LLC-1 tumors were allowed to grow for an additional 10 days. Measurement of tumor volume indicated that i.v. administration of GFP-expressing MSCs did not grow LLC-1 tumors in NPRA-KO mice, and that NPRA-KO tumors were substantially smaller than WT tumors (Supporting Information Fig. S4A, S4B). Reduced tumor burden in NPRA-KO mice correlated with significantly decreased numbers of LLC-1 cells in these tumors, as judged from immunostaining of the tumor sections for a tumor marker, CEA (Supporting Information Fig. S4C, S4D). These results suggest that LLC-1 cells are less able to graft or failed to proliferate in NPRA-KO mice.

On day 21, the tumors were excised, sectioned, and examined for CD31, SMA and F4/80 that represent endothelial progenitors, CAFs, and macrophages, respectively. Significantly increased numbers of NPRA, CD31, SMA, and macrophages were found in the TME of WT compared to NPRA-KO (Fig. 4A). To further confirm these findings, the sections were immunostained for additional characteristic cell markers such as S100A4 and vimentin representing CAFs, and vWF representing ECs. Results showed that WT tumor sections express significantly higher levels of CAFs (vimentin and S100A4) and ECs (vWF) compared to NPRA-KO tumor sections (Supporting Information Fig. S5A). Moreover, tumors from WT mice also exhibited more GFP +ve cells than NPRA-KO tumors, which colocalized with endothelial progenitor marker (CD31), EC marker (vWF) and CAF markers (SMA, vimentin and S100A4) (Fig. 4B; Supporting Information Fig. S5B). Numbers of GFP-expressing F4/80 +ve macrophages were not different in the two groups. These results collectively suggest that host NPRA signaling is essential for attracting stem cells to the TME that differentiated to EPCs, ECs, and CAFs. Consistent with this conclusion, dual immunostaining of the WT tumor sections for NPRA and CD31, SMA or GFP-expressing MSCs show that NPRA is also expressed in the TME and colocalized with endothelial progenitors, CAFs and recruited MSCs, as suggested by the Mander's overlap coefficient (R = 0.860, 0.613, and 0.813, respectively) (Supporting Information Fig. S6).

thumbnail image

Figure 4. NPRA signaling controls MSC recruitment to the tumor microenvironment. LLC-1 were injected s.c. into the flanks of WT and NPRA-KO mice. GFP expressing MSCs were injected i.v. on day 10, and tumors were extracted on day 21. (A): NPRA, endothelial progenitor cells (CD31), cancer associated fibroblasts (SMA), and macrophages (F4/80) were detected by immunohistochemistry. Five fields (×400) from each group were counted for indicated cell types (red) and the groups were statistically compared with Student's t test. (B): GFP-expressing MSCs (green) and CD31, SMA and F4/80 (red) in the tumor microenvironment were stained. MSCs that were differentiated into the above cell types appeared orange to yellow. Five fields (×400) from each group were counted for GFP−expressing MSCs (green), particular cell types (red) and differentiated stem cells (yellow) with ImageJ and the groups were statistically compared with Student's t test. *, p < 0.05; **, p < 0.01; ***, p < 0.001. Scale bar = 20 μm. Abbreviations: GFP, green fluorescent protein; MSCs, mesenchymal stem cells; NPRA, natriuretic peptide receptor A; KO, knock out; SMA, smooth muscle actin; WT, wild type.

Download figure to PowerPoint

Coimplantation of MSCs and Tumor Cells Increased Tumor Growth in NPRA-KO Mice

To determine whether the impaired angiogenesis seen in NPRA-KO mice was strictly due to lack of cell recruitment, we cocultured aortic rings from NPRA-KO mice with MSCs and examined aortic sprouting in these cultures. When MSCs were placed on the Matrigel containing aortic rings in vitro, the capillary growth increased significantly only in the NPRA-KO (Fig. 5C, 5D) mice aortic rings. These results indicate that the reduced vasculogenic potential of NPRA-KO aortic rings might be in part due to lack of angiogenic factors that can be restored by circulating MSCs.

thumbnail image

Figure 5. MSCs restore vasculogenesis in NPRA-KO aorta. Aortas from WT and NPRA-KO mice were cut into 1–2 mm rings and cultured in Matrigel for 7 days with or without MSCs (mag. ×40; n = 6 rings per group) (A, C). The areas of capillary growth were quantitated using ImageJ. Areas were expressed as means ± SEM (B, D) and statistically compared with Student's t test. *, p< 0.05. Abbreviations: MSCs, mesenchymal stem cells; KO, knock out; WT, wild type.

Download figure to PowerPoint

To confirm the role of NPRA signaling in cell recruitment in vivo, we examined tumor burden and angiogenesis in WT and NPRA-KO mice coinjected s.c. with LLC-1 cells and MSCs into the flanks. Control groups were injected with only LLC-1 cells. The tumors were monitored for 3 weeks, and on day 21 tumors were excised and TES aliquots were examined. The results showed that injection of LLC-1 cells in NPRA-KO mice failed to increase tumor burden, however, coinjection of MSCs with LLC-1 cells increased tumor burden in these mice as judged by increase in tumor weight and volume (Fig. 6A, 6B). Increased tumor burden in these mice correlated with increased VEGF expression in the TES (Fig. 6C), and expression of endothelial progenitors (CD31), endothelium (Tie2) and CAFs (SMA) in the TME of NPRA-KO mice (Fig. 6D, 6E). These results underscore the importance of NPRA signaling in cell migration and recruitment of BM progenitors. However, they did not clarify whether coinjected MSCs have directly differentiated or have attracted other cells to the microenvironment without undergoing differentiation.

thumbnail image

Figure 6. Coimplantation of MSCs with tumor cells restored tumorigenesis in NPRA-KO mice. LLC-1 cells were coinjected s.c. with or without MSCs into the flanks of WT and NPRA-KO mice. After 3 weeks, the tumors were measured, excised, and weighed. (A, B): Tumor dimensions and tumor weights are shown as means ± SEM (n = 4; *, p< 0.05). (C): Tumor explant supernatants from the tumors were analyzed for VEGF by ELISA. (D, E): Frozen tumor sections were stained for Tie2, CD31, and SMA (D) and fluorescent cell numbers were counted using ImageJ (E), expressed as means ± SEM (n = 10 fields per group; one field per section, two sections per mouse) and statistically compared by one-way ANOVA. *, p < 0.05; **, p < 0.01; ****, p < 0.0001. (F): Differentiation of MSCs in NPRA-KO tumor microenvironment. LLC-1 cells were injected into the flanks of NPRA-KO mice with GFP-expressing MSCs. After 3 weeks, the tumors were excised, sectioned, immunostained, and analyzed for differentiation of GFP-MSCs to cancer associated fibroblasts (SMA, S100A4), and endothelial progenitors (CD31). R is Mander's overlap coefficient. An R value >0.5 indicates that the GFP-expressing MSCs have substantially differentiated into the indicated cells. Scale bar = 20 μm. Abbreviations: GFP, green fluorescent protein; LLC, Lewis lung carcinoma; MSC, mesenchymal stem cells; KO, knock out; SMA, smooth muscle actin; VEGF, vascular endothelial growth factor; WT, wild type.

Download figure to PowerPoint

To further evaluate the differentiation potential of MSCs in the TME of NPRA-KO mice, GFP-expressing MSCs were coinjected with LLC-1 cells s.c. into the flanks of NPRA-KO mice, and tumors were monitored for 3 weeks. The presence of MSCs increased the tumor size (Supporting Information Fig. S7). Tumor sections were evaluated for coexpression of GFP and the corresponding cell markers that indicate stem cell differentiation. Substantial numbers of GFP-MSCs were found differentiated into CAFs (SMA +ve and S100A4 +ve cells) and endothelial progenitors (CD31 +ve), as suggested by the Mander's overlap coefficient (R = 0.803, 0.766, and 0.796, respectively) (Fig. 6F). These findings suggest that the influence of NPRA signaling on tumor angiogenesis is limited by the cell recruitment process and is not due to the lack of MSC differentiation.

NPRA Signaling Modulates CXCR4-SDF1α Expression in the TME

Rapid tumor growth causes various regions of the tumor to exhibit hypoxia and ischemia [29]. CAFs in these regions secrete SDF1, which interacts with its receptor CXCR4, and attracts CXCR4 +ve stem cells and immune cells to facilitate tumor angiogenesis [16, 21]. CXCR4 activation also results in increased VEGF production, which contributes to increased angiogenesis [21]. Since cell migration is deficient in NPRA-KO mice, we examined whether NPRA signaling influences CXCR4/SDF1α expression during in vivo tumorigenesis. Analysis of tumor sections by immunohistochemistry indicated that NPRA-KO tumors showed substantially reduced CXCR4 (Fig. 7A) and low to nonexistent SDF1α expression (Fig. 7B). However, coimplantation of MSCs with LLC-1 cells, which increased tumorigenesis in NPRA-KO mice also induced SDF1α expression. These findings indicate that NPRA signaling influences cell migration though modulation of SDF1α.

thumbnail image

Figure 7. MSCs recruited into NPRA-KO tumors correlated with restoration of CXCR4 and SDF1α. Tumors excised 21 days after LLC-1 cell injection from WT and NPRA-KO mice were cut into 10 μM sections, stained for CXCR4 (A) and SDF1α (B), and examined under the fluorescence microscope (×400; n = 5 fields per group; 1 field per section per mouse). Numbers of fluorescent cells were counted in ImageJ, expressed as means ± SEM, and statistically compared by one-way ANOVA. Tumor explant supernatants from the tumors were analyzed for SDF1α by ELISA (C). *, p < 0.05; **, p < 0.01; ****, p < 0.0001. Scale bar = 20 μm. Abbreviations: CXCR4, chemokine (C-X-C motif) receptor 4; SDF1α, stromal cell derived factor 1α; KO, knock out; WT, wild type.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

Tumor growth beyond 1 mm3 in the body requires angiogenesis, which involves formation of new blood vessels from pre-existing vessels [30, 10] and, therefore, antiangiogenic strategies constitute one of the major cancer treatment modalities [31]. The results of the studies reported herein provide evidence that NPRA signaling plays a pivotal role in regulating inflammation-induced angiogenesis required for tumor growth. Further, NPRA signaling regulates the inflammation in the TME by specifically controlling recruitment of progenitor cells that are critical for tumor growth in part via modulation of the SDF1-CXCR4 axis.

One of the major findings of our studies is that NPRA signaling plays a significant role in inflammation-induced angiogenesis as assayed by endothelial sprouting. Three lines of evidence support this notion. First, while LPS promoted sprouting, NPRA signaling inhibitors such as anantin inhibited sprouting. Second, aortic cultures from NPRA-KO mice show reduced potential to differentiate to CD31 +ve endothelial progenitors cells and to grow capillary sprouts relative to WT mice. Third, ECs from NPRA-KO mice formed fewer and less-defined capillary tubes. In addition to the sprout assay, we investigated the molecular correlates of NPRA signaling-induced angiogenesis. Culture supernatants from the NPRA-KO aortic cells were found to produce significantly less VEGF than WT, which agrees with previous reports [4, 32]. Furthermore, our results show that NPRA-KO sprouts have decreased numbers of CD31 +ve and CXCR4 +ve cells suggesting that the lack of differentiation potential of NPRA-KO mice is in part due to reduced expression of CXCR4. Together, these findings demonstrate the importance of NPRA signaling in regulating inflammation-induced angiogenesis by modulating recruitment of the CXCR4 expressing EPCs.

Another major finding of our studies relates to the role of NPRA signaling in promoting tumor stroma interaction. A comparison of the cellular components of TME of WT and NPRA-KO mice showed that the TME of NPRA-KO mice had a significantly reduced accumulation of CAFs, ECs, and macrophages compared to WT TME. A lack of these cells in TME accounts for the impaired tumor growth in NPRA-KO mice. To further dissect the mechanism of how NPRA signaling affects TME and angiogenesis, we examined whether lack of NPRA expression in TME, as in KO mice, induced a defect in migration of EPCs and other cells to the tumor site. Consistent with our hypothesis, when administered i.v., very few GFP-expressing MSCs migrated to tumors in NPRA-KO mice compared to WT mice, which suggested that NPRA-KO mice are indeed defective in cell migration. Surprisingly, coinjection of MSCs with tumor cells, which bypasses the need for cell recruitment, increased expression of progenitors in the TME of NPRA-KO mice. In addition, MSC coinjection accentuated tumor growth and blood vessel formation. This is consistent with other studies, where MSCs have been shown to support tumor growth after being integrated into the TME [33–43]. The reason for the increase of angiogenesis and tumor growth in the NPRA-KO mice is unknown at present, but there are two possibilities. First, the coinjected MSCs could have differentiated directly into various cells. Further evaluation indicated that stem cell differentiation was unaffected in the absence of NPRA signaling. Second, stem cells might have attracted other cells to the tumor milieu through secretion of cytokines/chemokines. It should be noted, however, that these two possibilities are not mutually exclusive. Irrespective of the precise mechanism of the lack of angiogenesis in NPRA-KO mice, our results have demonstrated that migration of progenitor cells to the tumor site is a key determinant in NPRA regulation of angiogenesis.

Another mechanistically important finding of our studies is that CXCR4 expression and SDF1α secretion were dependent on NPRA signaling. Tumor-induced inflammatory chemokines released into the circulation mobilize stem cells from bone marrow to the site of inflammation [44], and SDF1α has been shown to play a crucial role in activation of MSCs [45]. Our findings revealed that the absence of NPRA signaling caused significantly fewer MSCs to migrate to the TME. Further examination showed that NPRA signaling correlated expression of CXCR4 and SDF1α secretion. At present we do not know if NPRA signaling directly regulates CXCR4 and/or SDF1α and what the potential downstream signaling pathway(s) comprise. These data do, however, emphasize the influence of NPRA on cell recruitment to the TME through the SDF1α/CXCR4 axis.

Of particular interest is the finding that MSCs produce VEGF and SDF1α, which is consistent with several previous reports that have shown that MSCs are source of several growth factors and cytokines including VEGF [46, 47] and SDF1α [48, 49], that increase the immune tolerance in the TME. For instance, it has been reported that immunosuppressive myeloid derived suppressor cells (MDSCs) migrate to TME and promote tumor progression [50] and blocking VEGF [51] or SDF1α/CXCR4 [52] signaling hinders MDSC migration to TME. In the context of this study, we have found that tumors in WT mice contain significantly higher number of Gr1 +ve MDSCs compared to in NPRA-KO mice (Nagaraj et al., unpublished observation) suggesting that higher levels of VEGF and SDF1α produced in the TME of WT mice might promote increased tumor growth. However, it is to be noted that in our MSC studies we have no direct evidence that VEGF and SDF1α can compensate for MSC coimplantation. Our findings are merely correlative and they do not imply that VEGF and SDF1α are responsible for the increased tumor growth in NPRA-KO mice. Whether VEGF and SDF1α produced by MSCs can at least partly play a role in promoting tumors directly or indirectly by allowing migration of MDSCs remain to be elucidated.

CONCLUSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

In conclusion, NPRA signaling plays a critical role in stem cell progenitor recruitment to form a reactive stroma that promotes tumor angiogenesis in part by activating SDF1α/CXCR4 axis and increasing VEGF secretion.

To our knowledge, this is the first study implicating NPRA signaling in MSC migration and tumor stroma interaction. Consequently, manipulation of NPRA signaling in the tumor stroma might be an effective approach to control angiogenesis and hence tumorigenesis. Since MSCs also migrate to other sites of injury [53] and inflammation [44, 54], our study has broad implications and could explain the general anti-inflammatory activity of NPRA inhibitors and antagonists.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

This work is supported by 5R01CA152005 grants from National Institute of Health to S.M. and S.S.M., and 09BW-08 Florida Biomedical Research grants awarded to S.M. We thank Dr. Gary Hellermann for a critical reading of the manuscript.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
sc-12-0923_sm_SupplFigure1.tiff2702KSupplementary Figure S1: NPRA as a biomarker for melanoma, colon and pancreatic cancer. Human melanoma, colon and pancreatic cancer tissue microarray (TMA), prepared in the histology laboratory of the Moffitt Cancer Center Tissue Core Facility was used to test for expression of NPRA. The TMA slide was stained using a Ventana Discovery XT automated system (Ventana Medical Systems, Tucson, AZ), according to the manufacturer's protocol, as described 2. The TMA slide was scored for intensity and cellularity by an expert pathologist, as described 2. The product of the intensity and percentage scores was used as the final score. The final score was classified as: 0, negative; 1-3, weak; 4-6, moderate; and 7-9, strong. A median analysis of NPRA scores and the frequency in each disease group of having a score at or below the median was performed. (A) Scores for NPRA in different stages of melanoma are shown. These include, normal skin (N) (n=21), NEVI (compound Nevi, junctional nevi, intradermal nevi) (n=21), Clark's atypical dysplastic nevi (DN) (n=37). (B) Scores for each sample of different stages of Colon cancer are shown. These include, normal (N) (n = 24); adenoma (ADE) (n = 24) and adenocarcinoma (ADC) (n = 78). (C) Scores for each sample of different stages of pancreatic cancer are shown. These include, normal pancreatic duct (N) (n = 54); Intraductal carcinoma IDC (n = 68) and neuroendocrine carcinoma (NC) (n = 158). The bar represents the mean sample score for each category.
sc-12-0923_sm_SupplFigure2.tif812KSupplementary Figure S2: Control antibody staining. A) Specificity of antibodies was optimized by immunohistochemical staining of WT LLC-1 tumor sections using corresponding host- and isotype-matched control antibodies. Control and test sections were represented in the same row. Rb IgG: rabbit IgG. B) LLC-1 cells and GFP expressing MSC were stained with antibodies for CEA (upper panel) and GFP (lower panel). Scale bar in all the images is 20 μm.
sc-12-0923_sm_SupplFigure3.tiff2702KSupplementary Figure S3: Lung vasculature in NPRA-KO mice. Lungs isolated from WT and NPRA-KO mice were sectioned and stained for vWF. Cells expressing vWF were counted and compared between WT and KO sections. n = 4 mice/group, 1 section/mouse. Fluorescent cells were counted, expressed as means ± SEM, and compared with unpaired student t-test. *p < 0.05.
sc-12-0923_sm_SupplFigure4.tiff2702KSupplementary Figure S4: LLC-1 cells were injected s.c. into the flanks of WT and NPRA-KO mice. GFP-BM MSCs were injected i.v. on day 10 and tumors were extracted on day 21. Tumor dimensions (A) and weights (B) were measured and statistically compared with Student's t-test. (C-D) Tumor sections were stained for carcinoembryonic antigen (CEA) antigen and examined with a fluorescence microscope (400× magnification; n = 4 mice/group; 1 section per mouse). Fluorescent cells were counted, expressed as means ± SEM, and compared with unpaired student t-test. **P < 0.01. Scale bar in the images is 20 μm.
sc-12-0923_sm_SupplFigure5A.tiff2702KSupplementary Figure S5: After LLC-1 cells were injected s.c. into the flanks of WT and NPRA-KO mice, GFP-BM MSCs were injected i.v. on day 10 followed by tumor extraction on day 21. (A) Tumor sections were stained for CAFs (S100A4, Vimentin) and endothelial cells (vWF). Cells from WT and KO tumor sections were quantified (n = 3 mice; one field/ section/mice) and compared with student's t-test. (B) Tumor sections were dual stained for GFP and characteristic markers such as S100A4, Vimentin, and vWF. The dual stained cells were counted (n = 3 mice; one field/ section/mouse) and the WT and KO sections were compared with student's t-test. **p < 0.01, ***p < 0.001, ****p < 0.0001
sc-12-0923_sm_SupplFigure5B.tiff2702KSupplementary Figure S5: After LLC-1 cells were injected s.c. into the flanks of WT and NPRA-KO mice, GFP-BM MSCs were injected i.v. on day 10 followed by tumor extraction on day 21. (A) Tumor sections were stained for CAFs (S100A4, Vimentin) and endothelial cells (vWF). Cells from WT and KO tumor sections were quantified (n = 3 mice; one field/ section/mice) and compared with student's t-test. (B) Tumor sections were dual stained for GFP and characteristic markers such as S100A4, Vimentin, and vWF. The dual stained cells were counted (n = 3 mice; one field/ section/mouse) and the WT and KO sections were compared with student's t-test. **p < 0.01, ***p < 0.001, ****p < 0.0001
sc-12-0923_sm_SupplFigure6.tiff2702KSupplementary Figure S6: LLC-1 tumors were generated by s.c. injection of LLC-1 cells followed after 10 days by i.v. injection of GFP-MSC. The tumors were excized on day 21. The tumor sections (n = 3 mice; one field/ section/mouse) were dual stained for NPRA expression and for specific cells, (a) endothelial cells (CD31), (b) CAFs (SMA), and (c) MSCs (GFP). Manders coefficient was calculated in imageJ to assess the dual stain expression. Scale bar in the images is 20 μm.
sc-12-0923_sm_SupplFigure7.tiff2702KSupplementary Figure S7: LLC-1 cells with or without MSCs were injected s.c. into the flank region of NPRA-KO mice. The tumors were measured at the end of 3 weeks. Tumor A) volumes and B) weights were represented as means ± SEM and compared with student t-test in Graphpad prism. L: LLC-1 cells; M: MSCs; **P ≤ 0.01.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.