Human Lung Spheroids as In Vitro Niches of Lung Progenitor Cells with Distinctive Paracrine and Plasticity Properties

Abstract

Basic and translational research on lung biology has discovered multiple progenitor cell types, specialized or facultative, responsible for turnover, renewal, and repair. Isolation of populations of resident lung progenitor cells (LPCs) has been described by multiple protocols, and some have been successfully applied to healthy human lung tissue. We aimed at understanding how different cell culture conditions may affect, in vitro, the phenotype of LPCs to create an ideal niche-like microenvironment. The influence of different substrates (i.e., fibronectin, gelatin, laminin) and the impact of a three-dimensional/two-dimensional (3D/2D) culture switch on the biology of LPCs isolated as lung spheroids (LSs) from normal adult human lung biopsy specimens were investigated. We applied a spheroid culture system as the selective/inductive step for progenitor cell culture, as described in many biological systems. The data showed a niche-like proepithelial microenvironment inside the LS, highly sensitive to the 3D culture system and significantly affecting the phenotype of adult LPCs more than culture substrate. LSs favor epithelial phenotypes and LPC maintenance and contain cells more responsive to specific commitment stimuli than 2D monolayer cultures, while secreting a distinctive set of paracrine factors. We have shown for the first time, to our knowledge, how culture as 3D LSs can affect LPC epithelial phenotype and produce strong paracrine signals with a distinctive secretomic profile compared with 2D monolayer conditions. These findings suggest novel approaches to maintain ex vivo LPCs for basic and translational studies. Stem Cells Translational Medicine 2017;6:767-777.

Keywords: Epithelial-to-mesenchymal transition; Lung stem cells; Pneumospheres; Stem cell niche; Three-dimensional culture.

Figure 1

Culture protocol and experimental timing of pneumosphere explant culture. (A): Schematic representation of cell culture protocol. (B): Representative phase‐contrast images of lung explant cultures, PSs, and PDCs on different coatings. Yield was analyzed by averaging the PS number per field (C) and the PS diameter (D), and was similar on the three coatings tested. Scale bars = 250 µm. Abbreviations: BEGM, bronchial/epithelial growth medium; CM, conditioned medium; FN, fibronectin; GEL, gelatin; LAM, laminin; PDC, pneumosphere‐derived cell; PS, pneumosphere.

Figure 2

Protein expression profile of pneumospheres (PSs) and pneumosphere‐derived cells (PDCs) in different culture conditions by immunofluorescence. (A–E): Representative confocal images of PSs. (A): Small, early‐forming PSs were highly positive for the stemness marker Oct4 and the proliferation marker ki67, with frequent double‐stained cells (white arrows). (B): Mature PSs remained positive for the transcription factor TTF‐1. (C): Few cells inside the PS were positive to VIM staining, particularly on the outer layers, but cells migrating out of the PS started expressing VIM as they attached as monolayers. (D): Aquaporin 5 was expressed by a scattered portion of cells inside the PS. (E): Cells on the outer layers of the PS were strongly positive to KRT5 and pro‐SFTPC staining, consistent with a phenotypical gradient. (F): Examples of clusters of KRT5+/SFTPC+ double‐positive cells just detached from a PS (dotted line) could be identified. Representative fluorescence images of PDCs are shown, which remained highly positive only for Oct4 (G), Vim (H), ki67 (I), and TTF‐1 (J) staining. All PS images were taken at the confocal plane of the core unless otherwise specified. Scale bars = 50 µm. Abbreviation: Vim, vimentin.

Figure 3

PDC phenotype profiling and comparative gene expression profile with pneumospheres (PSs). (A): PDC doubling times on different substrates. (B): Gene expression levels for multiple markers of interest were analyzed by real‐time polymerase chain reaction (PCR) and compared among different coatings. (C): Representative image of PDC clonogenesis experiments, and corresponding efficiency analysis of different coatings by semiautomatic identification and quantification of clones (white asterisks). (D): PDC immunophenotype was unaffected by the coating and was analyzed by cytofluorimetry for a panel of hematopoietic, vascular, and mesenchymal markers, supporting their stromal nature. (E): Gene expression levels were analyzed by real‐time PCR and compared between PSs and PDCs. Expression levels for genes of both undifferentiated and committed profiles were significantly upregulated in three‐dimensional (3D) PSs compared with two‐dimensional (2D) PDCs, suggesting an enhanced niche‐like microenvironment inside the PS. ∗, p < .05 versus plastic or 2D; #, p < .01 versus 2D. Abbreviations: AT1, alveolar type 1 pneumocyte; AT2, alveolar type 2 pneumocyte; EMT, epithelial to mesenchymal transition; Endo, endothelium; epith, epithelium; FN, fibronectin; GEL, gelatin; LAM, laminin.

Figure 4

Differential response of PSs and PDCs to BEGM culture. (A): PSs were responsive to 1 week of culture in BEGM medium, as shown by significant upregulation of multiple gene expression levels compared with control medium, whereas PDCs were mostly unresponsive. The y‐axes are plotted on the log scale. (B): As indicated by immunofluorescence staining after BEGM culture, vimentin expression in PSs was low under control conditions, and high expression levels of Oct4 were confirmed inside the PS. KRT18 expression was low but detectable, whereas pro‐SFTPC was expressed in outer cell layers on a thicker portion than under basic culture conditions. All PS images were taken at the confocal plane of the core, unless specified. Scale bars = 50 µm. ∗, p < .05; #, p < .01 versus the reference (two‐dimensional culture or control medium). Abbreviations: AT1, alveolar type 1 pneumocyte; AT2, alveolar type 2 pneumocyte; BEGM, bronchial/epithelial growth medium; CEM/PGM, complete explant medium/pneumosphere growth medium; EMT, epithelial to mesenchymal; Endo, endothelium; epith, epithelium; PDC, pneumosphere‐derived cell; PS, pneumosphere.

Figure 5

PS‐CM and PDC‐CM screening for paracrine and signaling molecules. (A): Representative profile of PS‐CM analyzed by protein array. A selection of growth factors, chemokines, and cytokines of interest is highlighted in the panel. A complete list of factors screened by the assay is presented in supplemental online Table 2. Differential analysis of conditioned media from PSs and PDCs revealed a distinctive modulation in the secretion profile of a subset of cytokines. PS/PDC ratios are plotted in (B), direct level comparison is shown in (C). Factors in plots are listed based on decreasing PS/PDC ratio. Abbreviations: O.D., optical density; PDC, pneumosphere‐derived cell; PDC‐CM, pneumosphere‐derived cell‐conditioned medium; PS, pneumosphere; PS‐CM, pneumosphere‐conditioned medium.