Hydrogel-Embedded Precision-Cut Lung Slices Model Lung Cancer Premalignancy Ex Vivo

Abstract

Lung cancer is the leading global cause of cancer-related deaths. Although smoking cessation is the best prevention, 50% of lung cancer diagnoses occur in people who have quit smoking. Research into treatment options for high-risk patients is constrained to rodent models, which are time-consuming, expensive, and require large cohorts. Embedding precision-cut lung slices (PCLS) within an engineered hydrogel and exposing this tissue to vinyl carbamate, a carcinogen from cigarette smoke, creates an in vitro model of lung cancer premalignancy. Hydrogel formulations are selected to promote early lung cancer cellular phenotypes and extend PCLS viability to six weeks. Hydrogel-embedded PCLS are exposed to vinyl carbamate, which induces adenocarcinoma in mice. Analysis of proliferation, gene expression, histology, tissue stiffness, and cellular content after six weeks reveals that vinyl carbamate induces premalignant lesions with a mixed adenoma/squamous phenotype. Putative chemoprevention agents diffuse through the hydrogel and induce tissue-level changes. The design parameters selected using murine tissue are validated with hydrogel-embedded human PCLS and results show increased proliferation and premalignant lesion gene expression patterns. This tissue-engineered model of human lung cancer premalignancy is the foundation for more sophisticated ex vivo models that enable the study of carcinogenesis and chemoprevention strategies.

Keywords: biomaterials; hydrogels; in vitro models; lung cancer; precision-cut lung slices; premalignancy; tissue engineering.

Figure 1:

PEG-NB hydrogels were tuned to recapitulate key features of the premalignant microenvironment. A) Schematic of the experimental design for generating tissue-engineered models of lung cancer premalignancy in hydrogel-embedded PCLS. B) The compositions of hydrogel formulations tested to support ex vivo premalignancy development. C) Rheology results for soft and stiff hydrogel formulations revealed that these materials mimic the mechanical properties of healthy (1–5 kPa) and diseased (>10 kPa) human lung tissue, respectively. Soft hydrogels had an elastic modulus of 5.05 ± 0.20 kPa while stiff have a modulus of 14.69 ± 1.78 kPa (N=3). D) Maximum intensity projection of a confocal image. Hydrogel-embedded murine PCLS showing lung tissue labeled with CellTracker (red) and hydrogel labeled with a maleimide-conjugated AlexaFluor488 (green). Scale bar, 200 μm.

Figure 2:

Dose testing showed exposure to 10 or 50 μg ml⁻¹ vinyl carbamate did not cause overt toxicity in hydrogel-embedded PCLS. A) Presto blue metabolic assessment of hydrogel-embedded PCLS displayed maintenance of viability over six weeks in culture (N=10–11). B) Representative maximum intensity projections from confocal imaging show live (green) and dead (red) staining of hydrogel-embedded PCLS at six weeks. Viability of tissue at the center of the 3D construct was maintained over time. Scale bar, 100 μm C) Quantification of live/dead staining at week six revealed that 50–65% of cells were still alive in all constructs with no statically significant differences (two-way ANOVA; Tukey Test. N=11). D) qPCR analysis showed upregulation of Cyp2e1 in response to exposure to the higher dose of vinyl carbamate (one-way ANOVA; Kruskal-Wallis test. N=3).

Figure 3:

Stiff hydrogels with a low concentration of collagen mimic peptide (GFOGER) provided an environment that supported lung cancer premalignancy in response to vinyl carbamate exposure. A) Representative confocal images (maximum intensity projections) of cellular proliferation within hydrogel-embedded PCLS after six weeks exposure to 50 μg ml⁻¹ vinyl carbamate. Actively proliferating cells incorporated EdU and were labeled with AlexaFluor488-azide (green). Total nuclei were stained with Hoechst (blue). B) Quantification of the EdU assay revealed increases in proliferation due to vinyl carbamate exposure (N=5–9; two-way ANOVA, Tukey Test). C) Expression analysis of genes altered by urethane exposure showed a pattern in stiff/low collagen hydrogel after vinyl carbamate that is similar to urethane-exposed murine lung (N=3; the scale bar is fold change in expression relative to vehicle control). D) A hydrogel formulation with an elastic modulus of 12.3 kPa and a CGFOGER concentration of 0.1 mM was identified as the best for maximizing proliferation, Ttf1 and Cox2 expression while minimizing Apc2 and Bnip2 expression.

Figure 4:

Lesions formed in the tissue-engineered models recapitulated characteristic markers of premalignancy. A) Representative H&E images of hydrogel-embedded PCLS following six weeks of exposure to 50 μg ml⁻¹ vinyl carbamate in the best hydrogel formulation. Regions of macrophage infiltration (left, single-headed arrows), alveolar atypia (left, double-headed arrows), fibrosis (center), and bronchial dysplasia (right, single-headed arrows) were identified by a pathologist. Boxes indicate regions assessed by AFM. Scale bar, 50 μm. B) AFM analysis showed that both in vivo and ex vivo premalignant lesions had high stiffness relative to normal lung tissue. C) Topography maps of the 50 × 50 μm regions identified by boxes in panel A showed variable tissue heights up to 10 μm. D) Stiffness maps of the same regions showed the variability in relative elastic modulus between ex vivo premalignant lesions. E) qPCR analysis of six genes implicated in lung carcinogenesis revealed gene expression patterns consistent with premalignancy observed in vivo, as compared to vehicle treated controls (two-tailed T-test; N=3–5).

Figure 5:

Exposure to vinyl carbamate alters immune and cancer cell markers in hydrogel-embedded PCLS. A) Representative immunofluorescent images stained for CD3 (T-cells), CD68 (macrophages), and Lectin (tissue structure) showed an accumulation of macrophages in vinyl carbamate-exposed, hydrogel-embedded PCLS. B) Quantification of positive CD68 signal, normalized to tissue area (combined DAPI and Lectin signal) per slice and presented as fold change over the average of PBS controls. C) Representative immunofluorescent images stained for Ki67 (proliferation), CD68 (macrophages), and Lectin (tissue structure) demonstrated active proliferation of macrophages. D) Representative immunofluorescent images stained for CYP2E1 (vinyl carbamate metabolism), KRT5 (squamous carcinoma), and TTF1 (adenocarcinoma) revealed regions of mixed premalignant phenotype in vinyl carbamate-exposed, hydrogel-embedded PCLS. E) Quantification of positive Krt5, Cyp2e1, and Ttf1 stain normalized to cellularity (DAPI signal) per lesion (or equivalent size normal tissue region) and presented as fold change over the average of the PBS controls. All scale bars, 100 μm.

Figure 6:

Chemoprevention agents diffused through hydrogels and act on hydrogel-embedded PCLS. A) A Presto Blue assay for metabolic activity monitored viability of hydrogel-embedded PCLS at day 0, 3, and 7 during treatment with a vehicle control or 10 μM iloprost. Data are shown relative to day 0 (dotted line). There were no statistical differences between vehicle and treatment (two-way ANOVA; Tukey Test; N = 8–10). B) Expression of Pparγ was measured by qPCR at day 7 and presented relative to vehicle control (dotted line; two-tailed T-test, N=4). C) Activity of PPARγ was increased on day 7 (two-tailed T-test; N=5). D) A Presto Blue assay measured viability of PCLS at day 0, 3, and 7 during treatment with vehicle or 3 μg ml⁻¹ curcumin. Data are shown relative to day 0 (dotted line). There were no statistical differences between vehicle and treatment (two-way ANOVA; Tukey Test; N=12.) E) Expression levels of curcumin responsive genes were measured by qPCR at day 7 and shown relative to vehicle control (dotted line; two-tailed T-test, N=4–5).

Figure 7.

Human, hydrogel-embedded PCLS exposed to vinyl carbamate developed premalignant phenotypic changes over six weeks in culture. A) A Presto Blue assay measured viability of hydrogel-embedded or unembedded PCLS weekly through day 42 during exposure to vehicle or vinyl carbamate (50 μg ml⁻¹). Unembedded PCLS showed reduced viability relative to embedded counterparts at multiple timepoints (two-way ANOVA, Tukey test; N=6–18. Data are shown relative to day 0 (dotted line) B) EdU assay demonstrated increased proliferation in human PCLS as a response to vinyl carbamate exposure (two-tailed T-test, N=3–12; Scale bar, 200 μm). C) Expression of genes expected to be altered by vinyl carbamate was measured by qPCR at day 42, showing consistency with premalignant phenotype (two-tailed T-test; N=3–5).