Role of Homeobox A1 in Airway Epithelial Generation from Human Airway Basal Cells

Abstract

Airway basal cells from chronic obstructive pulmonary disease patients show a reduction in HOXA1 expression and generate an abnormal airway epithelium. Because the specific role of HOXA1 in airway basal cells is not known, we investigated the contribution of HOXA1 in the generation of the airway epithelium, which depends on basal cell proliferation, polarization, and differentiation. Airway stem cells were transduced with an inducible HOXA1 shRNA lentivector to knock down HOXA1 in either proliferating cells or100% confluent cells. The bronchial epithelium expresses HOXA1 near the basement membrane, likely representing basal cells. HOXA1 knockdown in proliferating basal cells attenuated cell proliferation. HOXA1 knockdown in confluent monolayers of basal cells generated an abnormal airway epithelium characterized by goblet cell hyperplasia and an inflammatory phenotype. Compared to the control, HOXA1 knockdown cells showed a decrease in transepithelial resistance, localization of occludin and E-cadherin to the intercellular junctions, reduced expression of occludin but not E-cadherin, and increased expression of TNF-α. Blocking TNF-α increased the expression of occludin in HOXA1 K/D cells. Based on these results, we conclude that HOXA1 plays an important role in cell proliferation, polarization, and differentiation, which are essential steps in airway epithelial generation. Additionally, HOXA1 may regulate occludin expression by inhibiting TNF-α expression.

Keywords: airway epithelial repair; cell polarization; cell proliferation; goblet cell metaplasia; mucociliary-differentiated cell cultures; occludin.

Figure 1

HOXA1 is expressed in the bronchial epithelium, and expression is reduced in patients with COPD. Green, HOXA1; blue, nuclei. (A,B) Lung sections from normal and COPD subjects stained with HOXA1 antibody. (C) Normal lung section stained with isotype control IgG. The images are representative of lung sections from 3 healthy non-smokers.

Figure 2

HOXA1 K/D airway basal cells show attenuated proliferation compared to control cells. (A,B) Representative images of control and HOXA1 K/D cells, respectively. (C) Data represent the range with the median from two independent experiments performed in triplicate. (D,E) Data represent the size of the spheres averaged from two independent experiments performed in triplicate (t test). Two to three random fields per culture were used to determine the number and size of the spheres. (F) Spheres were dissociated from each well, and the total number of cells was estimated. Data represent the range with the median from two independent experiments performed in triplicate (Mann–Whitney test).

Figure 3

Mucociliary-differentiated airway epithelial cell cultures generated from HOXA1 K/D basal cells show more goblet cells and less ciliated cells than the control. (A,B) Confocal immunofluorescence images showing ciliated cells (green), goblet cells (red), and nuclei (blue) in control and HOXA1 cultures, respectively. Images are representative of 3 independent experiments. (C) The number of goblet cells, ciliated cells (green), and other cell types (DAPI) was counted per 100 µm², and the data are presented as the range with the median.

Figure 4

HOXA1 K/D cell cultures show higher levels of pro-inflammatory cytokines than control cell cultures. Cytokine levels were measured in the basolateral medium from the mucociliary-differentiated cultures through ELISA. (A) IL-8, (B) IL-6, and (C) TNF-α. Results represent the mean ± S.D. calculated from 3 independent experiments (n = 3; unpaired t test).

Figure 5

HOXA1 K/D airway basal cells show a defect in tight junction formation. (A) The TER was measured, and the data represent the range with the median from 3 independent experiments performed in quadruplicate (n = 12, Mann–Whitney test, p ≤ 0.05, different from control). (B,C) Cells were fixed, blocked with BSA, and incubated with antibodies to occludin and E-cadherin. The bound antibodies were detected using antirabbit IgG conjugated with Alexa Fluor-594 (occludin) and antimouse IgG conjugated with Alexa Fluor 488 (E-cadherin). Cell cultures were counterstained with DAPI (blue) to detect nuclei and imaged under a confocal microscope. Images are representative of 3 independent experiments.

Figure 6

Quantitation of signal intensity of E-cadherin and occludin in polarized airway epithelial cells. Polarized cell cultures immuno-stained with occludin and E-cadherin were imaged under confocal microscopy, and signal intensities were measured as pixels in 10–12 random fields per Transwell; data are presented as the median with the range (n = 3, Mann–Whitney test).

Figure 7

Expression of occludin is decreased in HOXA1 K/O cells at the protein level. (A,C) Western blot showing the expression of E-cadherin and occludin, respectively. (B,D) The band densities were quantified using ImageJ and expressed as a fold change over GAPDH. Data in B and D are presented as the mean ± S.D. calculated from 3 independent experiments (n = 3, unpaired t test).

Figure 8

Inhibition of TNF-α improves occludin expression in HOXA1 K/D cell cultures. (A) The TNF-α level was measured in the basolateral medium of the control and HOXA1 K/D cells cultured at ALI for two weeks. (B) TNF-α mRNA levels were measured in HOXA1 K/D cell cultures treated with 2 µM TNF-α antagonist III or vehicle, and the results represent TNF-α expression levels normalized to GAPDH. (C) Western blot analysis of the total protein isolated from HOXA1 K/D cell cultures treated with 2 µM TNF-α antagonist III or vehicle. Data in A and B are presented as the mean ± S.D. calculated from 3 independent experiments (n = 3, unpaired t test).