Neuronal pentraxin 2 supports clear cell renal cell carcinoma by activating the AMPA-selective glutamate receptor-4

Abstract

Clear cell renal cell carcinoma (ccRCC) is the most common subtype of kidney cancer and has the highest propensity to manifest as metastatic disease. Recent characterizations of the genetic signature of ccRCC have revealed several factors correlated with tumor cell migration and invasion; however, the specific events driving malignancy are not well defined. Furthermore, there remains a lack of targeted therapies that result in long-term, sustainable response in patients with metastatic disease. We show here that neuronal pentraxin 2 (NPTX2) is overexpressed specifically in ccRCC primary tumors and metastases, and that it contributes to tumor cell viability and promotes cell migration through its interaction with the α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor subunit GluR4. We propose NPTX2 as a novel molecular target for therapy for patients with ccRCC diagnosed with or at risk of developing metastatic disease.

Figure 1. NPTX2 expression profile in clear cell Renal Cell Carcinoma

(A) TMA IHC of patient ccRCC vs. matched normal tissue for NPTX2 expression in stage I, II, III, IV (primary), and IV (metastatic) (normal n=44, 32, 35, 7, and 6 and tumor n= 41, 26, 33, 10, and 17 respectively). Cytoplasmic and membranous staining pattern observed. H-score ± standard deviation from the mean is shown. (B) Western blot of protein lysates prepared from NRE cells and ccRCC cell lines for NPTX2 expression. Protein expression level quantitation is normalized to β-actin, and total human brain tissue lysate was used as a positive control. (C) NPTX2 knockdown was evaluated in A498, KIJ265T, and Caki2 ccRCC cell lines infected with the sh1316 lentiviral construct as compared to NT controls via QPCR. (D) 7 day proliferation assay of ccRCC sh1316 clones vs. NT controls. (E) Cell death of A498, KIJ265T, and Caki2 NT vs. sh1316 cell populations analyzed via flow cytometry. (F) Western blot for total and cleaved PARP in NT vs. NPTX2 knockdown cells. Cells collected for cell death analysis (panels E and F) correspond to day 7 of proliferation assay (panel D).

Figure 2. NPTX2 modulates actin cytoskeletal remodeling and promotes invasion

(A) Phase-contrast microscopy of empty vector vs. NPTX2 transfected (+NPTX2) RWV366T and Caki1 cells, taken at 20x magnification. Arrows are used to highlight select membranous protrusions. IF of (B) empty vector vs. NPTX2 (HA tagged) transfected (+NPTX2) RWV366T cells and (C) A498 cells stained for NPTX2 (left panel) and VASP (center panel). Images are merged in the right panel. 60x magnification of regions highlighted in 20x images are shown. (D) Pathway signature of EMT/cell migration in high-NPTX2 expressing tumors identified using Ingenuity Software analysis. Genes highlighted in orange are significantly upregulated in high-NPTX2 expressing tumors (P<0.05). QPCR for EMT-associated genes in NT vs. sh1316 NPTX2 knockdown (E) A498, (F) KIJ265T, and (G) Caki2 cells. Invasion assay of (H) empty vector vs. NPTX2 over-expressing Caki1 and RWV366T cells and (I) KIJ265T and A498 NT vs. sh1316 NPTX2 knockdown cells. Experiments were performed in triplicate, and representative images are displayed. Results are quantitated as invading cells per visual field.

Figure 3. GluR4 is overexpressed in ccRCC

(A) TMA IHC of patient ccRCC vs. matched normal tissue for NPTXR expression in stage I, II, III, IV (primary), and IV (metastatic) (normal n=12, 35, 34, 8, and 7 and tumor n= 13, 29, 34, 8, and 19 respectively). Human cerebellum tissue used as a positive control for NPTXR expression. (B) TMA IHC of patient ccRCC vs. matched normal tissue for GluR4 expression in stage I, II, III, IV (primary), and IV (metastatic) (normal n=45, 35, 38, 8, and 6 and tumor n= 39, 29, 34, 8, and 21 respectively). Cytoplasmic and membranous staining pattern observed. Human cerebellum tissue used as a positive control for GluR4 expression. H-score ± standard deviation from the mean is shown for NPTXR and GluR4 IHC analysis. (C) QPCR of NRE versus ccRCC cell lines for GluR1-4 (n=4 for both NRE and tumor samples). Tumor transcript expression is normalized to average NRE transcript expression. (D) Western blot of protein lysates prepared from NRE cells and ccRCC cell lines for GluR4 expression. Protein expression level quantitation is normalized to β-actin, and total human brain tissue lysate was used as a positive control.

Figure 4. GluR4 binds NPTX2 and promotes ccRCC cell viability

(A) Confocal immunofluorescence of non-permeabilized Caki2 NT, sh1316, and sh1676 clones for NPTX2 binding to the cell membrane. Each cell is outlined, and cross sections are displayed in the upper and right panels for each image. Relative fluorescence corresponding to NPTX2 expression is quantitated. IP of KIJ265T and Caki2 cells transfected with both HA epitope tagged human NPTX2 and Flag epitope tagged human GluR4 (HA-Flag) versus empty vector (EV) for (B) HA (NPTX2) mediated Flag (GluR4) pulldown and (C) Flag (GluR4) mediated HA (NPTX2) pulldown. WB of total lysate (Total) confirms expression of epitope tagged proteins. (D) QPCR for GluR4 expression in A498, KIJ265T, and Caki2 NT vs. sh1676 GluR4 knockdown cell populations. (E) 7 day proliferation assay of A498, KIJ265T, and Caki2 NT vs. sh1676 clones. (F) Cell death of A498, KIJ265T, and Caki2 NT vs. sh1676 cell populations analyzed via flow cytometry. (G) Western blot for total and cleaved PARP in A498, KIJ265T, and Caki2 NT vs. GluR4 knockdown cells. Cells collected for cell death analysis (panels F and G) correspond to day 7 of proliferation assay (panel E).

Figure 5. NPTX2-GluR4 mediates influx of intracellular calcium in ccRCC cells

(A) Proliferation of ccRCC cells in response to CFM-2 treatment (AMPA receptor antagonist) at day 7. (B) Effects of CFM-2 on A498 intracellular calcium levels using the cell permeable fluorescently labeled calcium indicator Calcium Green™-1, AM. Results are calculated as change in fluorescence over time (min). (*) Denotes significant change in fluorescence as compared to DMSO control per time point. (C) Western blot of KIJ265T and A498 cells treated with DMSO vs. CFM-2 (10µM, 1 hour) for phosphorylation of CAMK1(T177) and AKT(S473). (D) Western blot of R-NPTX2 (1ng/µL) treated Caki1 cells for phosphorylation of CAMK1(T177) and AKT(S473) treated with DMSO vs. CFM-2 (10µM) over time (min). (E) Invasion assay of DMSO vs. CFM-2 (10µM) treated KIJ265T and A498 cells. Experiments were performed in triplicate, and representative images are displayed. Results are quantitated as invading cells per visual field.