Patho-transcriptomic analysis of invasive mucinous adenocarcinoma of the lung (IMA): comparison with lung adenocarcinoma with signet ring cell features (SRCC)

Abstract

AbstractBackgroundInvasive mucinous adenocarcinoma (IMA) comprises ∼5% of lung adenocarcinoma. There is no effective therapy for IMA when surgical resection is not possible. IMA is sometimes confused with adenocarcinoma with signet ring cell features (SRCC) pathologically since both adenocarcinomas feature tumor cells with abundant intracellular mucin. The molecular mechanisms by which such mucin-producing lung adenocarcinomas develop remain unknown.MethodsUsing a Visium spatial transcriptomics approach, we analyzed IMA and compared it with SRCC patho-transcriptomically. Combining spatial transcriptomics data within vitrostudies using RNA-seq and ChIP-seq, we assessed downstream targets of transcription factors HNF4A and SPDEF that are highly expressed in IMA and/or SRCC.ResultsSpatial transcriptomics analysis indicated that there are 6 distinct cell clusters in IMA and SRCC. Notably, two clusters (C1 and C3) of mucinous tumor cells exist in both adenocarcinomas albeit at a different ratio. Importantly, a portion of genes (e.g.,NKX2-1,GKN1,HNF4AandFOXA3) are distinctly expressed while some mucous-related genes (e.g.,SPDEFandFOXA2) are expressed in both adenocarcinomas. We determined that HNF4A inducesMUC3A/BandTM4SF4and that BI 6015, an HNF4A antagonist, suppressed the growth of IMA cells. Using mutant SPDEF that is associated with COVID-19, we also determined that an intact DNA-binding domain of SPDEF is required for SPDEF-mediated induction of mucin genes (MUC5AC,MUC5BandAGR2). Additionally, we found that XMU-MP-1, a SPDEF inhibitor, suppressed the growth of IMA cells.ConclusionThese results revealed that IMA and SRCC contain heterogenous tumor cell types, some of which are targetable.