Integrating single-nucleus barcoding with spatial transcriptomics via Stamp-seq to reveal immunotherapy response-enhancing functional modules in NSCLC

Deciphering the spatial organization of cell states is fundamental for understanding development, tissue homeostasis and disease. Emerging advances in spatial transcriptomic profiling techniques allow transcript localization but face limitations in unambiguous cell state assignments due to cellular boundary inference, low gene detection and prohibitive cost. Here, a method, Stamp-seq, is developed that leverages custom-fabricated high-density DNA sequencing chips to label single nuclei with restriction enzyme-cleavable spatial barcodes. Stamp-seq spatial barcodes are distributed at a density of 1.6 μm on the chip, allowing for single physical cell resolution with precise subtype classification and spatial mapping (with an average 4 μm localization error) and reduced cost. We utilize Stamp-seq to delineate chemoimmunotherapy-responsive cellular ecosystems in non-small cell lung carcinoma, including a distinct IGHG1⁺ plasma cell-enriched community. Through a novel application of Stamp-seq to spatially resolve BCR clonotypes, we elucidate the spatiotemporal trajectory of treatment-potentiating IGHG1⁺ plasma cells, which originate from tertiary lymphoid structures (TLSs) or the vasculature, migrate through antigen-presenting CAF (apCAF)-enriched survival niches, and ultimately contact tumor cells. We highlight the power of spatial cellular subtyping and molecular tracking using Stamp-seq and suggest that the IGHG1⁺ plasma cell niche is a better prognostic biomarker for the chemoimmunotherapy response.