Differential expression analysis

To compare ATP6AP1 expression between BC and adjacent normal tissues, we used the Oncomine [37] (http://www.oncomine.org), GEPIA [38] (http://gepia.cancer-pku.cn/index.html) and HPA [39] (http://www.proteinatlas.org) databases. The Oncomine database draws relevant datasets directly from the Stanford Microarray Database, the National Center for Biotechnology Information Gene Expression Omnibus (GEO), published literature, etc. In Oncomine, mRNA data were selected with P = 0.05 and fold-change = 1.5 as the threshold values. The datasets in GEPIA are based on TCGA and GTEx, which contain normal tissue data for comparison.

The HPA database is a Swedish project to map all human proteins in cells, tissues and organs by integrating data from TCGA, HPA datasets, the GTEx consortium and recount2. We used the HPA database to assess ATP6AP1 protein levels in tumors and adjacent normal tissues. The antibody used to obtain the immunohistochemistry results was CAB015218 (Origene) at a dilution of 1:30. The immunohistochemistry results and antibody information are shown in Figure 1; however, due to the number of specimens in the database, we selected only a representative group for the figure, and placed the rest in the Supplementary Material.

GEO is a common functional genomics data repository. We analyzed expression profiles from GSE153277 [40] and GSE155241 [41]. GSE153277 contains nine samples from iAT2 cells infected with SARS-CoV-2 or a mock virus. GSE155241 contains data from hPSC-LO cells cultured with SARS-CoV-2 or a mock virus. Nine samples were selected from a total of 18. The GPL18573 Illumina NextSeq 500 (Homo sapiens) and GPL24676 Illumina NovaSeq 6000 (Homo sapiens) platforms were used to sequence the respective datasets. ATP6AP1 levels were compared between the control and SARS-CoV-2 groups using the Sangerbox tool, a free online platform for data analysis (http://www.sangerbox.com/tool).