Optimized isolation and expansion of human airway epithelial basal cells from endobronchial biopsy samples

Abstract Autologous airway epithelial cells have been used in clinical tissue‐engineered airway transplantation procedures with a view to assisting mucosal regeneration and restoring mucociliary escalator function. However, limited time is available for epithelial cell expansion due to the urgent nature of these interventions and slow epithelial regeneration has been observed in patients. Human airway epithelial cells can be expanded from small biopsies or brushings taken during bronchoscopy procedures, but the optimal mode of tissue acquisition from patients has not been investigated. Here, we compared endobronchial brushing and endobronchial biopsy samples in terms of their cell number and their ability to initiate basal epithelial stem cell cultures. We found that direct co‐culture of samples with 3T3‐J2 feeder cells in culture medium containing a Rho‐associated protein kinase inhibitor, Y‐27632, led to the selective expansion of greater numbers of basal epithelial stem cells during the critical early stages of culture than traditional techniques. Additionally, we established the benefit of initiating cell cultures from cell suspensions, either using brushing samples or through enzymatic digestion of biopsies, over explant culture. Primary epithelial cell cultures were initiated from endobronchial biopsy samples that had been cryopreserved before the initiation of cell cultures, suggesting that cryopreservation could eliminate the requirement for close proximity between the clinical facility in which biopsy samples are taken and the specialist laboratory in which epithelial cells are cultured. Overall, our results suggest ways to expedite epithelial cell preparation in future airway cell therapy or bioengineered airway transplantation procedures.


| INTRODUCTION
Airway tissue engineering has seen the development of cell-scaffold solutions for otherwise intractable human disease and has seen clinical translation in compassionate cases (Badylak, Weiss, Caplan, & Macchiarini, 2012). Preclinical (Crowley, Birchall, & Seifalian, 2015) and early clinical work (Hamilton et al., 2015) suggests the importance of epithelial restoration following airway transplantation to avoid complications, such as recurrent infection (Zhang, Fu, & Xu, 2015).
Clinical protocols have recognized the importance of epithelial replacement as bioengineered airway scaffolds have included cultured epithelial cells or explanted airway mucosa from recipient airways (Elliott et al., 2012). However, a standardized method to expand these cells is lacking, largely due to the varied clinical scenarios in which transplantation is indicated. Such technology would also be invaluable to inform the development of 'standalone' respiratory mucosal epithelial replacement treatments for a range of diseases. We and others have recently characterized an improved culture methodology (Chapman, Liu, Meyers, Schlegel, & McBride, 2010;Liu et al., 2012;Suprynowicz et al., 2012) for human airway basal cells, the stem/progenitor cells of the human upper airways (Hogan et al., 2014), by co-culturing epithelial cells with mitotically inactivated 3T3-J2 feeder cells in medium containing the Rho-associated protein kinase inhibitor Y-27632 (3T3+Y) (Butler et al., 2016;Reynolds et al., 2016). This method has clear advantages over the time-consuming derivation of airway epithelial cells from pluripotent stem cells and conventional cell culture using bronchial epithelial growth medium (BEGM) for basal cell expansion, owing to its capability to expand autologous primary cells from living patients in meaningful numbers. However, the optimal method to isolate autologous epithelial cells from patient biopsy samples remains unclear.
Previously, this has been achieved by explant culture of endobronchial biopsy samples in BEGM (Butler et al., 2016), but we reasoned that initiation of cultures would be improved using the 3T3+Y protocol.
In vitro expansion of human airway epithelial cells has been reported from both endobronchial brushings (Kelsen, Mardini, Zhou, Benovic, & Higgins, 1992) and endobronchial biopsies, either as explants (de Jong, van Sterkenburg, Kempenaar, Dijkman, & Ponec, 1993) or digested to obtain a cell suspension (Goulet et al., 1996). (e) Cell counts after trypsinization of epithelial outgrowths at day 12 of culture. A statistical analysis was performed using a two-way ANOVA with Bonferroni post-test; *p < 0.05; ****p < 0.0001; n = 8-14 biopsy samples within each condition (with a minimum of four donors sampled per group). (f) Flow cytometric analysis of cell outgrowths for basal epithelial cell markers cytokeratin 5 (CK5) and integrin α6. The percentage represents the mean ± standard error of the mean; n = 3-7 biopsy samples within each condition (with a minimum of two donors sampled per group) Initially we sought to compare two alternative modes of tissue acquisition to derive autologous airway cells from patients: endobronchial biopsy and endobronchial brushing samples (Figure 1a). Individuals consented to either endobronchial biopsies or endobronchial brushings, so we compared these modes of acquisition at the population level. We found a trend towards there being fewer cells in biopsy samples than in brushings, although this difference was not statistically significant ( Figure 1b). This could be explained by the inefficiency of the digestion protocol used to isolate cells from biopsy tissue.
Next, we investigated methods to derive epithelial cell cultures from these endobronchial samples. We compared the expansion of epithelial cells from brushings and biopsies seeded directly in co-culture as well as biopsies that were digested to a single cell suspension prior to seeding (Figure 1c). Cells were seeded both in direct 3T3+Y co-culture and in medium conditioned by 3T3-J2 feeder cells (CM+Y), as it has been previously reported that secreted factors mediate the effects of co-culture and that direct contact with feeder cells is not required After trypsinization and re-seeding (that is, at 'passage 1'), we found that basal cells derived from brushings and biopsies behaved Cell counts after trypsinization of epithelial outgrowths after 7 days of subculture. A statistical analysis was performed using a two-way ANOVA with Bonferroni post-test; *p < 0.05; **p = 0.01, ***p < 0.001; n = 7-13 donor cell cultures within each condition (with a minimum of four donors sampled per group). (c) Flow cytometric analysis of passage one epithelial cells for basal epithelial cell markers cytokeratin 5 (CK5) and integrin α6. The percentage represents the mean ± standard error of the mean; n = 3-9 biopsy samples from between two and four donors. (d) Representative images and quantification of colony-forming assays to investigate the nature of 3T3-J2 feeder cell support of human airway epithelial cells. Epithelial cells were grown in direct co-culture with 3T3-J2 cells, in indirect co-culture with 3T3-J2 cells (separated by a transwell) or in 3T3-J2-conditioned medium. A statistical analysis was performed using a Kruskal-Wallis test; *p < 0.05; colony-forming assays were performed in triplicate using cells derived from three donors similarly (Figure 2a), with comparable numbers generated in either 3T3 +Y co-culture or in CM+Y, although consistently fewer cells were expanded over the 7-day culture period in CM+Y (Figure 2b). Again, CK5/integrin α6 co-expression was demonstrated by flow cytometry (Figure 2c).
Next, we used low-density colony formation assays to investigate the cause of the decreased numbers of basal cells expanded in CM+Y.
Fewer epithelial colonies were generated in CM+Y than in direct 3T3 +Y co-culture but separation of epithelial cells from feeder cells by a transwell membrane allowed colony formation that was comparable with direct co-culture (Figure 2d). This indicated that, although direct cell-cell contact between epithelial cells and feeder cells was dispensable for the enhancement of epithelial cell expansion, continuous production of secreted factors by feeder cells was important for epithelial support and could not be reproduced by feeding cells with 3T3-J2conditioned medium three times per week.
Finally, as organoid derivation was recently reported from cryopreserved primary tumours (Walsh, Cook, Sanders, Arteaga, & Skala, 2016), we investigated whether airway epithelial cell cultures could be initiated from cryopreserved patient samples. We generated cell cultures from cryopreserved samples with an efficiency of 89% for explant biopsies, 70% for biopsies digested prior to cryopreservation, 67% for biopsies digested after cryopreservation and 100% for brushings (Figure 3a, b). Again, we found that plating a single cell suspension, either by digesting a biopsy before or after cryopreservation or by using a brush biopsy, generated the highest number of cells ( Figure 3c). To ensure that cryopreserved biopsy-derived basal cells maintained their capacity for ciliated differentiation, we expanded cells for one further passage and used a three-dimensional tracheosphere assay (Danahay et al., 2015) (Figure 3d).
To conclude, direct explant expansion of human airway epithelial cells in 3T3+Y culture conditions improves the early stages of culture compared with BEGM (88% culture success, approximately 400 000 cells after 12 days vs. 50% culture success, <250 000 cells after 2 weeks; Butler et al., 2016). However, the cell yield of explant biopsies is inefficient compared with initiating cultures from a cell suspension, either by digestion of biopsy samples (88% culture success, approximately 4 × 10 6 cells after 12 days) or by collecting cells by brush biopsy (94% culture success, approximately 1.8 × 10 6 cells after 12 days). We estimate that isolation of cells from a cell suspension in 3T3+Y could reduce the time required for epithelial cell expansion to 3-4 weeks in a case where only basal cells were to be transplanted on an adult tracheal scaffold. Colony formation assays suggested that airway epithelial cells required a continual supply of feeder products that was not recreated by feeding three times per week with 3T3-J2-conditioned medium: when cell-cell contact was prevented by adding 3T3-J2 feeder cells on a transwell membrane, epithelial colony growth was comparable with growth in direct coculture. Notably, our study revealed that patient tissue can be cryopreserved prior to the initiation of cell cultures. Although cryopreservation did slow the initiation of epithelial cell cultures, this finding could remove the necessity for specialist clinical and research facilities to be in close proximity in future airway and mucosal tissue-engineering clinical procedures. Current protocols limit the number of sites capable of performing autologous airway cell facilities, but samples could be cryopreserved for transportation, expanding the potential application of these techniques. Overall, our findings suggest methods to facilitate cell preparation in future bioengineered airway transplantation procedures.