Silencing PCBP2 normalizes desmoplastic stroma and improves the antitumor activity of chemotherapy in pancreatic cancer

Abstract

Rationale: Dense desmoplastic stroma is a fundamental characteristic of pancreatic ductal adenocarcinoma (PDAC) and comprises up to 80% of the tumor mass. Type I collagen is the major component of the extracellular matrix (ECM), which acts as a barrier to impede the delivery of drugs into the tumor microenvironment. While the strategy to deplete PDAC stroma has failed in clinical trials, normalization of the stroma to allow chemotherapy to kill the tumor cells in the "nest" could be a promising strategy for PDAC therapy. We hypothesize that silencing the poly(rC)-binding protein 2 (αCP2, encoded by the PCBP2 gene) leads to the destabilization and normalization of type I collagen in the PDAC stroma. Methods: We develop a micro-flow mixing method to fabricate a peptide-based core-stabilized PCBP2 siRNA nanocomplex to reverse the accumulation of type I collagen in PDAC tumor stroma. Various in vitro studies were performed to evaluate the silencing activity, cellular uptake, serum stability, and tumor penetration of the PCBP2 siRNA nanocomplex. We also investigated the penetration of small molecules in stroma-rich pancreatic cancer spheroids after the treatment with the PCBP2 siRNA nanocomplex. The anti-tumor activity of the PCBP2 siRNA nanocomplex and its combination with gemcitabine was evaluated in an orthotopic stroma-rich pancreatic cancer mouse model. Results: Silencing the PCBP2 gene using siRNA reverses the accumulation of type I collagen in human pancreatic stellate cells (PSCs) and mouse NIH 3T3 fibroblast cells. The siRNA nanocomplex significantly reduces ECM production and enhances drug penetration through desmoplastic tumor stroma. The combination of gemcitabine with the PCBP2 siRNA nanocomplex markedly suppresses the tumor progression in a desmoplastic PDAC orthotopic mouse model. Conclusion: This approach provides a new therapeutic avenue to improve the antitumor efficacy of PDAC therapies by normalizing tumor stroma using the PCBP2 siRNA nanocomplex.

Keywords: PCBP2; nanocomplex; pancreatic cancer; siRNA; tumor stroma.

Scheme I

Schematics of the core-stabilized PCBP2 siRNA nanocomplex and its combination with gemcitabine for PDAC therapy. (A) Schematics of the cholesterol-peptides/PCBP2 siRNA nanocomplexes. Three types of cholesterol-peptides are used to spontaneously fold into micelle-like particles, followed by condensation with PCBP2 siRNAs using a micro-flow mixing method to form the siRNA nanocomplex. (B) After internalization and release from the endosome in the activated PSCs, the released PCBP2 siRNA blocks the expression of ⍺CP2 and subsequently reverses the accumulation of type I collagen in the PDAC stroma, leading to enhanced antitumor efficiency of gemcitabine.

Figure 1

Silencing activity of the PCBP2 siRNA in human PSCs and NIH 3T3 mouse fibroblasts. (A) mRNA expression of the PCBP2 gene in human PSCs after incubation with the PCBP2 siRNA condensed with Lipofectamine RNAiMAX for 24 h. Representative western blot images of type I collagen and ⍺CP2 in human PSCs (B) and NIH 3T3 mouse fibroblasts (C) after incubation with the PCBP2 siRNA condensed with Lipofectamine RNAiMAX. Quantitative analysis of ⍺CP2 (D) and type I collagen (E) protein expression using ImageJ. (F) mRNA expression of the PCBP2 gene in human PSCs after incubation with the CP/PCBP2 siRNA nanocomplex at different N/P ratios for 24 h. (G) mRNA expression of the PCBP2 gene in NIH 3T3 cells after incubation with the CP/PCBP2 siRNA nanocomplexes at the N/P ratio 5:1 for 6 and 24 h. All results are presented as the mean ± SD (n = 3). For western blot, three independent experiments were conducted for quantitative analysis. For RT-PCR, total RNA from three independent samples was isolated to measure the silencing effect. (*P < 0.05, **P < 0.01)

Figure 2

Characterization of the cholesterol-peptides/PCBP2 siRNA nanocomplex. (A) CMC values of CP, CPC and CCP micelles. (B) Depiction of the micro-flow mixing device for siRNA nanocomplex preparation. (C) Particle size and (D) PDI of CP, CPC and CCP siRNA nanocomplexes prepared with the micro-flow mixing method at different flow rates. Representative size distribution (E) and zeta potential (F) of CP, CPC and CCP siRNA nanocomplexes prepared with the micro-flow mixing method at a 0.05 mL/min flow rate. (G) Representative TEM images of the siRNA nanocomplexes prepared with the micro-flow mixing method at 0.05 mL/min. The top scale bars represent 1 µm, and the bottom scale bars represent 500 nm. (H) Serum stability of the siRNA nanocomplex in 50% mouse serum. All results are presented as the mean ± SD (n = 3 independent experiments).

Figure 3

Silencing activity of the PCBP2 siRNA nanocomplexes in NIH 3T3 mouse fibroblasts. (A) mRNA expression of the PCBP2 gene in NIH 3T3 mouse fibroblasts after incubation with CP, CPC and CCP/ PCBP2 siRNA nanocomplexes for 6 h. (B) Representative immunostaining images of ⍺CP2 and type I collagen in NIH 3T3 cells after incubation with the CP, CPC and CCP/PCBP2 siRNA nanocomplexes. Scale bars represent 100 μm. Quantitative analysis of ⍺CP2 (C) and type I collagen (D) expressions using image J. The expressions were normalized to scrambled siRNA treated group. All results are presented as the mean ± SD (n = 3 independent experiments). For RT-PCR, total RNA from three independent samples was isolated to measure the silencing effect. For immunostaining, three independent experiments were conducted for each group for quantitative analysis. (*P < 0.05, **P < 0.01)

Figure 4

Cellular uptake of the cholesterol-peptide/siRNA nanocomplexes in NIH3T3 mouse fibroblasts. Flow cytometry analysis of NIH 3T3 mouse fibroblasts after incubation with free Cy5-siRNA or Cy5-siRNA nanocomplexes for 2 h (A) and 4 h (C). The fluorescence intensity of labeled NIH 3T3 cells at 2 h (B) and 4 h (D). (E) Representative confocal images of NIH 3T3 mouse fibroblasts after incubation with free Cy5-siRNA or Cy5-siRNA nanocomplexes for 4 h. Scale bars represent as 20 μm. All results are presented as the mean ± SD (n = 3 independent experiments). (**P < 0.01)

Figure 5

Penetration in stroma-rich 3D pancreatic cancer spheroids. The stroma-rich 3D pancreatic cancer spheroids were composed of PANC-1 tumor cells and NIH 3T3 fibroblasts. (A) Representative scanning images of the tumor spheroids with a z-stack of 25 μm after 2 h incubation with free Cy5-siRNA or Cy5-siRNA nanocomplexes. Scale bars represent as 100 μm. (B) Mean fluorescence intensity of free Cy5-siRNA and Cy5-siRNA nanocomplexes in the scanning images vs. the distance from the spheroid periphery. All results are presented as the mean ± SD (n = 3 independent spheroids).

Figure 6

The CCP/PCBP2 siRNA nanocomplex enhances the penetration of small molecules in stroma-rich pancreatic cancer spheroids and cytotoxicity studies. The stroma-rich 3D pancreatic tumor spheroids were composed of PANC-1 tumor cells and NIH 3T3 fibroblasts. Representative z-stack confocal images of the stroma-rich tumor spheroids incubated with Hoechst 33258 for 2 h (A) and 4 h (B). Scale bars represent as 200 µm. Mean fluorescence intensity of the z-stack scanning confocal images vs. the distance from the periphery of the spheroids for 2 h (C) and 4 h (D). (E) Representative z-stack images of the stroma rich tumor spheroids at 100 µm from the periphery. Scale bars represent as 200 µm. (F) Quantification of the penetration depth of Hoechst 33258. (G) Percentage of penetration depth of Hoechst 33258 based on the tumor spheroid's size. All results are presented as the mean ± SD (n = 3 independent spheroids). (*P < 0.05, **P < 0.01) Cytotoxicity of NIH 3T3 (H) and PANC-1 (I) cells in 2-D cell culture after incubation with the siRNA nanocomplexes for 24 h. The results are presented as the mean ± SD (n = 5 independent samples).

Figure 7

Antitumor therapeutic efficacy of the CCP/PCBP2 siRNA nanocomplex combined with gemcitabine in an orthotopic mouse model of desmoplastic pancreatic tumor. (A) Scheme of combination treatment. The desmoplastic orthotopic model was established by co-inoculation of PANC-1/Luc tumor cells with NIH 3T3 fibroblasts into the tail of the pancreas in nude mice on Day 0. The mice received saline, CCP/scrambled siRNA nanocomplex, and CCP/PCBP2 siRNA nanocomplex via tail vein on Days 10, 12, 17, and 19. The mice also received saline or gemcitabine on Days 14, 21, and 28. The mice were euthanized on Day 34 post-implantation. (B) In vivo bioluminescence images of mice on Days 19, 26 and 33. (C) Tumor growth curve was determined by the bioluminescence intensity. (D) Images of tumors with spleens. (E) The weight of tumors with spleens. The results are presented as the mean ± SD (n = 8 mice per group). *P < 0.05, **P < 0.01).

Figure 8

Histological assessment of tumor tissues. TUNEL assay (A), Ki67 (B), PCNA (C) immunohistochemistry (IHC) staining and Picrosirius Red (D) staining. The scale bar represents 200 μm. Quantitative analysis of apoptotic cells areas in TUNEL assay (E) and Picrosirius Red stained areas (F) using Image J. The data are presented as the mean ± SD (n = 4 tumors per group, 4 images from each tumor were taken for quantitative analysis). (G) Western blot analysis of ⍺CP2 in tumor tissues. (H) Quantitative analysis of the normalized ⍺CP2 expression. The results are presented as the mean ± SD (n = 3 tumors per group). *P < 0.05; **P < 0.01