Development and Characterization of an In Vitro Model for Radiation-Induced Fibrosis

Abstract

Radiation-induced fibrosis (RIF) is a major side effect of radiotherapy in cancer patients with no effective therapeutic options. RIF involves excess deposition and aberrant remodeling of the extracellular matrix (ECM) leading to stiffness in tissues and organ failure. Development of preclinical models of RIF is crucial to elucidate the molecular mechanisms regulating fibrosis and to develop therapeutic approaches. In addition to radiation, the main molecular perpetrators of fibrotic reactions are cytokines, including transforming growth factor-β (TGF-β). We hypothesized that human oral fibroblasts would develop an in vitro fibrotic reaction in response to radiation and TGF-β. We demonstrate here that fibroblasts exposed to radiation followed by TGF-β exhibit a fibrotic phenotype with increased collagen deposition, cell proliferation, migration and invasion. In this in vitro model of RIF (RIF_iv), the early biological processes involved in fibrosis are demonstrated, along with increased levels of several molecules including collagen 1α1, collagen XIα1, integrin-α2 and cyclin D1 mRNA in irradiated cells. A clinically relevant antifibrotic agent, pentoxifylline, and a curcumin analogue both mitigated collagen deposition in irradiated fibroblast cultures. In summary, we have established an in vitro model for RIF that facilitates the elucidation of molecular mechanisms in radiation-induced fibrosis and the development of effective therapeutic approaches.

Figure 1. Radiation and TGF-β induce collagen deposition in oral fibroblasts

(A) Oral fibroblasts exposed to radiation (3 Gy) and/or treated with TGF-β (100 ng/ml) exhibited collagen deposition. Cells were stained with Direct Red 80 to detect collagen (200× magnification). (B) Cumulative quantitative analyses of collagen deposition by OD assessment of Direct Red 80. Graph depicts three replicated experiments plated in duplicate, with fold change of each experiment indexed to vehicle control treated group, two-tailed Mann-Whitney tests used to calculate p-values, error bars represent +/− SEM.

Figure 2. Radiation and TGF-β enhance fibroblast proliferation and migration

(A) Fibroblasts treated with radiation (3 Gy) and/or TGF-β (100 ng/ml) over 72 h demonstrate an increase in relative cell number (p<0.0001, graph depicts three replicate experiments plated in triplicate), (B) expression of cyclin D1 as determined by RT-PCR and (C) quantified using densitometry analyses of cyclin D1 expression (p=0.05, graph depicts imageJ densitometry analyses of three PCR analyses, and indexed to untreated group). Fibroblasts were assessed for their ability to (D) migrate through a Boyden chamber and (E) invade a Matrigel matrix. Migration and invasion graphs depict three replicate experiments plated in triplicate normalized to parallel proliferation, and indexed to untreated group. Two-tailed Mann-Whitney tests used to calculate p-values, error bars on all graphs represent +/− SEM.

Figure 3. Radiation and TGF-β enhance anchorage independent growth and collagen deposition

(A) Fibroblasts formed spheres under non-adherent conditions (400× magnification). (B) The number of spheres was counted and (C) the diameter determined using Image-J software and data presented as fold change indexed to average diameter of controls. (D) Collagen deposition was assessed by measuring the OD of disassociated spheroids stained with Direct Red 80. Graphs depict three replicate experiments, plated in duplicate. Two-tailed Mann-Whitney tests used to calculate pvalues, error bars represent +/− SEM.

Figure 4. RNA-Seq analysis demonstrates molecular mechanisms of fibrosis

(A) A heat-map representing the log values of differentially expressed genes relative to the control demonstrates treatment-related changes in gene expression. Differentially expressed genes with an absolute fold change of ≥ 1.5 and q-value (false discovery rate) ≤ 0.05 were mapped. The expression data are hierarchically clustered in rows (distance metric: Euclidean; linkage method: Ward. (B) The network depicts Ingenuity® Knowledge Base pathway analyses of dominant genes with a fold change of 1.5 and p-value ≤ 0.05.

Figure 5. Radiation and TGF-β modulate gene expression in oral fibroblasts

RT-PCR followed by densitometry quantification demonstrate that radiation (3 Gy) combined with TGF-β (100 ng/mL) increases levels of (A, B) collagen XIα1, (C, D) collagen 1α1 and (E, F) integrin-α2, while significantly decreasing the levels of (G, H) HGF. Gene expression was normalized to β-actin levels, graphs depict three independent experiments and are indexed to control group, p-values determined using t-test with confirmation of normality using Shapiro-Wilk test, error bars represent +/− SEM.

Figure 6. Pentoxifylline and CDF decrease collagen deposition

Cells exposed to radiation (3 Gy) and TGF-β (100 ng/ml) were assessed for collagen deposition after treatment with antifibrotic agents. Cells in various treatment groups were stained with Direct Red 80, imaged under light microscopy and the OD quantified and graphed. Increasing concentrations of (A, B) pentoxifylline or (C, D) CDF reduced the level collagen deposition in the cultures. Collagen levels in each treatment arm were normalized to cell number. Graphs depict relative collagen deposition of three independent experiments plated in duplicate, two-tailed Mann-Whitney tests used to calculate p-values, error bars represent +/−SEM.