Airway Sampling Techniques for Single-Cell Profiling of Lung Biology

Rationale: Recent advances in single-cell technologies have begun to expand our understanding of lung development and disease. However, progress in animal models has not been matched in human studies. In part, this is due to the ability to isolate intact lung cells or nuclei for single-cell analysis as routinely. Samples of the airway lumen present challenges, including the presence of mucus and low viabilities of cells that can be isolated, and yet single-cell profiling of these samples could enable future studies in large cohorts or longitudinal studies. Therefore, we sought to develop protocols for isolating cells and generating high quality single-cell RNA sequencing (scRNA-seq) data from two airway lumen samples: induced sputum and endotracheal aspirates. Methods: We collected induced sputum from two healthy adults and tested seven different processing techniques to determine which conditions might influence scRNA-seq data quality. Additionally, we collected endotracheal aspirates from two intubated premature neonates (gestational age <28 weeks) and assessed the effects of time before cryopreservation on the quality of the data. Single-cell libraries were generated using a microfluidic platform and analysis was performed after quality filtering, including clustering, label transfer, differential expression, and trajectory analysis. Results: Despite lower-than-expected cellular yields for sputum, we were able to generate high quality data representing airway lumen biology. Notably, we observed that data quality was not improved by DTT or DNase treatment and that cells isolated from induced sputum could be cryopreserved before processing. Similarly, our protocol for processing endotracheal aspirates generated high quality data. We documented the effects on both cellular composition and transcriptional profiles of the isolated cells of up to 24-hour delays before cryopreservation. Interestingly, by integrating data from these two sample types (processed with minimal delay) we observe remarkably distinct states of myeloid cells. Myeloid cells from adult sputum are predominantly alveolar macrophages, while myeloid cells from aspirates of premature neonates represent a continuum of monocytes developing into interstitial macrophages and are largely devoid of alveolar macrophages. The dynamic nature of the cells in the neonate aspirate samples allowed us to chart the transcriptional trajectory of monocytes differentiating into interstitial macrophages. Conclusions: Our work demonstrates protocols for two airway lumen samples, complementing recently published studies. We find that tested variables for sputum protocols do not have large effects on data quality and ultimately by integrating the data from sputum and aspirates, we were able to observe important developmental steps in the neonatal lung.