Cold digestion provides greater yield and viability of AECs compared to hot digestion.
We used flow cytometry (Fig. 1a) to assess the yield of three different cell populations after a 1h 37°C dispase II/DNase I digestion of murine lungs. On average, we obtained 5.55 ± 0.05 x105 CD45−CD31−EpCAM+ AECs (7% of total cells), 7.35 ± 0.59 x106 CD45+ immune cells and 2.10 ± 0.03 x105 CD31+ endothelial cells (Fig. 1b).
We decided to investigate whether the use of cold digestion of murine lungs would increase the yield and/or viability of AECs. Using the cold digestion approach, the average number of isolated CD45−CD31−EpCAM+ AECs increased to 1.46 ± 0.15 x106, almost 3-fold the number obtained following hot digestion. Likewise, the number of CD31+ endothelial cells obtained with the cold digestion (4.27 ± 0.05 x105) was almost 2-fold higher than after hot digestion. Conversely, the number of CD45+ cells was similar between the two digestion techniques. This suggests that the cold digestion approach is beneficial for recovery of structural cells from the lung, while obtaining similar numbers of lung immune cells.
While several cell surface markers like CD4, CD8 or PD-1L are known to be sensitive to dispase II digestion25, we did not find any changes in expression levels of CD45, CD31 and EpCAM between the digestion methods as seen by comparable MFI (median) levels of each marker irrespective of digestion method (Fig. 1c). We also did not observe loss of several AECs-related cell surface markers such as major histocompatibility (MHC) I, MHC-II or CD24 (Supplementary Fig. 1, Table 1).
Following cold digestion, more AECs are viable compared to hot digestion.
Next, we compared the viability of isolated AECs between cold and hot lung digestions by flow cytometry, assessing the proportion of live cells within the CD45−CD31−EpCAM+ population in single cell suspension, with the gating strategy shown in Fig. 2a. We found that on average 83.99 ± 1.16% of CD45−CD31−EpCAM+ cells from cold lung digestions were alive, compared to 56.65 ± 3.16% after hot digestion, a 1.48-fold increase in viable AECs (Fig. 2b). Despite differences in viability of AECs, we did not observe changes in levels of Ldha (a marker of oxidative stress)26between digestion methods (Supplementary Fig. 2, Table 2). Given this increased viability, we hypothesised that MACS sorting, and subsequent seeding of AECs isolated by cold digestion would result in a greater number of basal AEC colonies in in vitro cultures, comparted to hot digestion. Performing double MACS sorting with initial CD45 and CD31 depletion followed by EpCAM positive selection, we recovered substantially more AECs following cold digestion (5.30 ± 0.45 x105 AECs) compared to hot digestion (2.97 ± 0.26 x105 AECs) which is equivalent to a 1.78-fold increase in number of viable AECs (Fig. 2c). The AECs recovered after MACS sorting were highly pure (Fig. 2d), with average purity of 96.35 ± 0.75% CD45−CD31−EpCAM+ cells of live cells. To confirm the airway epithelial identity of sorted CD45−CD31−EpCAM+ we used flow cytometry to confirm that these cells also express other epithelial markers. Virtually all CD45−CD31−EpCAM+ cells also expressed CD49f (integrin α6) 27,28 and CD24 29,30 (Fig. 2e), albeit as expected, at vastly varying levels.
Following cold digestion and MACS sorting, more basal cell colonies proliferate in vitro.
To investigate if the observed increase in AEC viability with cold digestion translated to greater recovery of airway basal cells in vitro, we seeded 2x105 MACS sorted CD45−CD31−EpCAM+ cells (98% purity) from hot or cold lung digestions into 24-well plates (workflow indicated in Fig. 3a). Cells were cultured for seven days in Promocell airway epithelial media supplemented with differentiation inhibitors and a Wnt pathway activator31–33. After seven days the cells were fixed and stained for markers of epithelial (E-cadherin) and basal cells (KRT5 and p63) (Table 3) 1,34. Then 25% of the surface area of each 24-well plate was imaged and the number of basal cell colonies was quantified, as well as the total surface area of KRT5+ basal cell colonies per well (Fig. 3b). KRT5+ colony morphology was visualized using phase-contrast microscopy (Fig. 3c). On average there were 21 ± 2.97 basal cell colonies after hot digestion and 34 ± 2.98 colonies after cold digestion per 52.6mm2, which equals to a 1.62-fold increase in colony counts. When we assessed the total surface area of KRT5+ colonies between the two digestion methods, we found a 2.3-fold increase in colony surface area with an average surface area occupied by KRT5+ colonies of 0.66 ± 0.13mm2 after cold digestions compared to 0.30 ± 0.07mm2 after hot digestions (Fig. 3d). Furthermore, there were very few E-cadherin negative non-epithelial cells in cultures irrespective of the digestion method employed, suggesting that the combination of cold digestion, MACS sorting, and appropriate media allows for propagating highly pure AECs.