Factors necessary to produce basoapical polarity in human glandular epithelium formed in conventional and high-throughput three-dimensional culture: example of the breast epithelium

Abstract

Background: Basoapical polarity in epithelia is critical for proper tissue function, and control of proliferation and survival. Cell culture models that recapitulate epithelial tissue architecture are invaluable to unravel developmental and disease mechanisms. Although factors important for the establishment of basal polarity have been identified, requirements for the formation of apical polarity in three-dimensional tissue structures have not been thoroughly investigated.

Results: We demonstrate that the human mammary epithelial cell line-3522 S1, provides a resilient model for studying the formation of basoapical polarity in glandular structures. Testing three-dimensional culture systems that differ in composition and origin of substrata reveals that apical polarity is more sensitive to culture conditions than basal polarity. Using a new high-throughput culture method that produces basoapical polarity in glandular structures without a gel coat, we show that basal polarity-mediated signaling and collagen IV are both necessary for the development of apical polarity.

Conclusion: These results provide new insights into the role of the basement membrane, and especially collagen IV, in the development of the apical pole, a critical element of the architecture of glandular epithelia. Also, the high-throughput culture method developed in this study should open new avenues for high-content screening of agents that act on mammary tissue homeostasis and thus, on architectural changes involved in cancer development.

Figure 1

The formation of differentiated glandular structures depends on the culture medium and cell line. A. Lumen delineated by ZO-1 as shown by immunohistochemistry (reddish-brown) in a mammary acinus and an alveolus from tissue biopsy sections. Hematoxylin staining indicates the position of the nuclei (blue). The drawing indicates the location of basal and apical poles in the glandular structure (nuclei are represented by filled black circles). B. Immunostaining for basal (laminin 332) [red], α6-integrin [α6-integrin; green]) and lateroapical (ZO-1 [green]) markers in acini formed by S1 cells in 3D culture. [Note: in vivo, the contact between luminal cells and the BM in the mammary gland is discontinuous and occurs where myoepithelial cells are not covering luminal epithelial cells [34]. In contrast, in vitro, contact with the BM occurs all along the basal surface of the luminal cells [35], therefore the α6-integrin staining is all around the acinus]. C. Microinjection of nondiffusible Rhodamine-Dextran (red) fluorescent dye in individual cells of acini formed by S1 cells in 3D culture. D. S1 and MCF10A cells were cultured for 15 days in 3D-Matrigel™ drip either in the H14 medium or the assay medium. Shown is the percentage of multicellular structures with apically polarized ZO-1 (Additional Table S1 with P values - see Additional file 1). Also shown is an example of ZO-1 staining in structures lacking apical polarity formed by MCF10A cells in 3D culture (arrowheads point to the basal location of ZO-1 at certain cell-cell junctions). ND = nondetected; size bar, 5 μm; *P < 0.05; **P < 0.01.

Figure 2

Apically localized tight junctions in mammary acini. A. Apical concentration of tight junction proteins PALS1 and PAR3 as shown by immunohistochemistry [reddish-brown; arrows] in acini from normal looking mammary tissue (from biopsy sections - left panel), and acini formed by S1 cells in 3D-Matrigel™ drip culture (center panel) as shown by immunofluorescence [green]. Diffuse distribution of PALS1 and PAR3 in multicellular structures formed by MCF10A cells in 3D culture (right panel). Nuclei are counterstained with 4', 6-diamidino-2-phenylindole (DAPI) (blue) in cell cultures and hematoxylin (blue) in IHC. Size bar, 5 μm. B. Electron micrograph of a section of an acinus formed by S1 cells in 3D culture. Insets show specific adhesion complexes. Size bar, 1 μm; A = apical; B = basal; L = lumen.

Figure 3

Distribution of apical polarity marker ZO-1 in acini depends on the substratum for 3D culture. A. Organization of the basoapical polarity axis in acini formed by S1 cells in 3D-Matrigel™ drip cultures as shown by immmunofluorescence staining of basal (collagen IV [green], α6-integrin [green]), lateral [beta-catenin [red] and desmoplakin 1,2 [green]), lateroapical [ZO-1 [red], and apical mucin-1 [green]) polarity markers. Nuclei are counterstained with DAPI (blue). Size bar, 5 μm. B. Histograms of the percentages of multicellular structures with apically localized ZO-1 (Additional Table S1 with P values - See Additional file 1) when S1 cells are cultured in the presence of chicken basal lamina (CBL) [36], PuraMatrix™, PuraMatrix™ and 20% laminin (PuraMatrix™ +20% L), PuraMatrix™ and 20% Matrigel™ (PuraMatrix™ + 20% MG), synthetic ECM (sECM-drip), synthetic ECM and 5% Matrigel™ drip (sECM + MG drip), monolayer on glass (Two-dimensional-2D), on glass + 5% Matrigel™ drip (High-throughput [HTP] drip) compared to their respective Matrigel™ controls. *P < 0.05; ***P < 0.001; ns = nonsignificant; ND = nondetected (no multicellular structures with apically located ZO-1).

Figure 4

HTP culture produces a basoapical polarity axis similar to that shown with classical 3D-Matrigel™ cultures. A. Organization of the basoapical polarity axis in 2D and HTP cultures of S1 cells as shown by immmunofluorescence staining of basal (collagen IV [green], α6-integrin [green]), lateral [beta-catenin [red] and desmoplakin 1,2 [red]), lateroapical [ZO-1 [green]), and apical (mucin-1 [green]) polarity markers. Nuclei are counterstained with DAPI (blue). Size bar, 5 μm. B. drawings of the different cell culture systems used in this study (2D monolayer; Matrigel™-embedded: cells totally embedded in the Matrigel™ gel; 3D-Matrigel™ drip: 5% Matrigel™ drip on Matrigel™ gel-coated surface; HTP: 5% Matrigel™ drip on glass surface). The gel is represented in red; soluble Matrigel™ dripped in the culture medium is represented by curvy lines.

Figure 5

Both basal polarity and collagen IV contribute to apical polarity in HTP cultures. A. Histogram of the percentages of acini formed by S1 cells (S1 acini) with lateroapical ZO-1 under different culture conditions (5% HTP, 10% HTP, control 5% 3D-Matrigel™ drip for eight and 10 days). B. Histograms of the percentages of S1 acini with lateroapical ZO-1, after treatment from day 5 to 8 with function blocking antibody for β1-integrin [β1-block] and β4-integrin [β4-block], and IgG in 5% HTP culture. C. Histograms of the percentages of S1 acini with lateroapical ZO-1 after culture in 2% HTP, 2% HTP + collagen IV (Col IV); 2% HTP + laminin 111 (lam 1); 2% HTP + collagen IV (Col IV) and laminin 111 (lam 1), during eight days of culture. *P < 0.05; **P < 0.01.