Nanoparticle-based therapeutic delivery of prohibitin to the colonic epithelial cells ameliorates acute murine colitis

Abstract

Background: Intestinal epithelial expression of antioxidants and nuclear factor kappa B (NF-κB) contribute to mucosal barrier integrity and epithelial homeostasis, two key events in the pathogenesis of inflammatory bowel disease (IBD). Genetic restoration of intestinal epithelial prohibitin 1 (PHB) levels during experimental colitis reduces the severity of disease through sustained epithelial antioxidant expression and reduced NF-κB activation. To determine the therapeutic potential of restoring epithelial PHB during experimental colitis in mice, we assessed two methods of PHB colonic mucosal delivery: adenovirus-directed administration by enema and poly(lactic acid) nanoparticle (NPs) delivery by gavage.

Methods: As a proof-of-principle to demonstrate the therapeutic efficacy of PHB, we utilized adenovirus-directed administration by enema. Second, we used NPs-based colonic delivery of biologically active PHB to demonstrate therapeutic use for human IBD. Colitis was induced by oral administration of dextran sodium sulfate (DSS) in water for 6-7 days. Wildtype mice receiving normal tap water served as controls.

Results: Both methods of delivery resulted in increased levels of PHB in the surface epithelial cells of the colon and reduced severity of DSS-induced colitis in mice as measured by body weight loss, clinical score, myeloperoxidase activity, proinflammatory cytokine expression, histological score, and protein carbonyl content.

Conclusions: This is the first study to show oral delivery of a biologically active protein by NPs encapsulated in hydrogel to the colon. Here we show that therapeutic delivery of PHB to the colon reduces the severity of DSS-induced colitis in mice. PHB may represent a novel therapeutic target in IBD.

Figure 1. GFP is transiently expressed in surface colonic epithelial cells 1 and 3 days post-infection in mice administered AdPHB

Mice were administered pAdTrack-CMV-PHB1-GFP (AdPHB) or pAdTrack-CMV-GFP (AdV) by enema. (A) Western blots showing GFP, PHB, and β-tubulin (loading control) expression from total protein isolated from colon mucosa collected day 1, 3 or 5 post-infection. (B–E) To show AdPHB localization, GFP fluorescence was visualized using fluorescent microscopy (B and D) and confocal microscopy (C and E) in colon tissue isolate 1 day post-infection. GFP expression (green) was increased in surface colonic epithelial cells in mice administered AdPHB (A–B) compared to mice administered PBS (C–D). Nuclei were stained with DAPI (blue). Bar = 20 µm in (B) and (D). Bar = 5 µm in (C) and (E).

Figure 2. Exogenous AdPHB-dependent inhibition of DSS-induced colitis

(A) Percent change in body weight during DSS colitis. *P < 0.05 vs. all other treatment groups. (B) Changes in clinical score. *P < 0.05 vs. AdV. (C) Colonic myeloperoxidase activity. *P < 0.05 vs. AdV. (D) Quantitative real-time PCR was used to quantify mRNA levels of the cytokines tumor necrosis factor α, interleukin-1β, and interferon γ. Values represent means ± SEM of 2 determinations per animal. *P < 0.05 vs. AdV. (E) Protein carbonyl content. *P < 0.05 vs. AdV. n = 5 animals per treatment.

Figure 3. Mice infected with AdPHB have reduced endoscopic and histologic damage induced by DSS compared to mice administered AdV

(A–D) Macroscopic inflammation was assessed using a mouse colonoscope. Photos were obtained on the day of sacrifice and an endoscopic score was calculated. (E–L) Representative photomicrographs of paraffin-embedded H&E-stained sections of distal colon. Original magnification 10× (E–H) and 40× (I–L). (M) Endoscopic score was derived as described in the Methods section. *P < 0.05 vs. AdV; n = 4 animals per treatment. (N) Histologic score of severity of inflammatory infiltrate, extent of crypt damage and ulceration used to generate a total histological score. **P < 0.01 vs. AdV; n = 5 animals per treatment.

Figure 4. Characterization of PHB-loaded nanoparticles (NPs)

(A) Size distribution of PHB-loaded NPs measured by light scattering. PHB-loaded NPs are 439.4 nm in diameter. (B) Scanning electron microscopy (SEM) picture of a suspension of NPs loaded with 1 mg/ml PHB. Bar = 667 nm. (C) Transmission electron microscopy (TEM) picture of a suspension of NPs loaded with 1 mg/ml PHB. Bar = 0.2 µm. (D) Cytotoxicity as measured by LDH test of a 1 mg/ml suspension of PHB-loaded NPs on Caco2-BBE cells for 30 hr and compared to BSA-loaded NPs and lipofectamine transfection as controls.

Figure 5. PHB-loaded NPs reduced TNFα-induced NF-κB activation in Caco2-BBE cells

(A) Confocal microscopy photograph of FITC-tagged PHB-loaded NPs (green) in polarized Caco2-BBE cell monolayers. Sections were counterstained with rhodamine/phalloidin and DAPI to visualize actin (red) and nuclei (blue), respectively. Untreated Caco2-BBE cells were used as a negative control. Bar = 20 µm. (B) TEM picture of Caco2-BBE cell exposed to 500 µg/ml PHB-loaded NPs for 48 hr. NPs are located in the cytosol of the cell. M, mitochondrion; N, nucleus; NPs, nanoparticle; V, villus. Bar = 1 µm. (C) Relative luciferase activity of Caco2-BBE cells transfected with the pNF-κB-Luc plasmid and exposed to BSA- (B) or PHB-loaded NPs for 72 hr. Cells were then treated with 10 ng/ml TNFα for 6 h. *P = 0.05 vs. vector; n = 3 per treatment. (D) EMSA showing binding of 10 µg nuclear proteins isolated from Caco2-BBE cells exposed to BSA- (B) or PHB-loaded (P) NPs for 72 hr and treated with 10 ng/ml TNFα for 30 min. Binding is to a consensus NF-κB site and is competed by unlabeled NF-κB oligonucleotides (cold). The addition of a p50 antibody caused a supershift.

Figure 6. Mice treated with PHB-loaded NPs exhibited less severe DSS-induced colitis

(A) FITC-PHB delivered by NPs is expressed in surface colonic epithelial cells. Colon was isolated and fluorescence was assessed by confocal microscopy. Bar = 5 µm. (B) Percent change in body weight during DSS colitis. *P < 0.05 vs. all other treatment groups. (C) Changes in clinical score. *P < 0.05 vs. AdV. (D) Colonic myeloperoxidase activity. *P < 0.05 vs. AdV. (E) Quantitative real-time PCR was used to quantify mRNA levels of the cytokines tumor necrosis factor α, interleukin-1β, and interferon γ. Values represent means ± SEM of 2 determinations per animal. *P < 0.05 vs. AdV. (F) Protein carbonyl content. *P < 0.05 vs. AdV. n ≥ 7 animals per treatment.

Figure 7. Mice treated with PHB-loaded NPs have reduced endoscopic and histologic damage induced by DSS

(A–C) Macroscopic inflammation was assessed using a mouse colonoscope. Photos were obtained on the day of sacrifice and an endoscopic score was calculated. (D–I) Representative photomicrographs of paraffin-embedded H&E-stained sections of distal colon. Original magnification 10× (D–F) and 40× (G–I). (J) Endoscopic score. *P < 0.05 vs. BSA-loaded NPs DSS; n = 4 animals per treatment. (N) Histologic score of severity of inflammatory infiltrate, extent of crypt damage and ulceration used to generate a total histological score. *P < 0.05 vs. BSA-loaded NPs DSS; n ≥ 7 animals per treatment.