Spatial Transcriptomics Uncovers the Hybrid Molecular Identity, Ciliated Phenotype, and Immune Signature of Adenomyosis Lesions

Adenomyosis is a common gynaecological condition characterised by ectopic endometrial tissue within the myometrium, causing debilitating symptoms with limited treatment options. Despite its prevalence, the pathogenesis of adenomyosis remains poorly understood, hindering the development of targeted therapies for true disease modification rather than symptom management. The objective of this study was to elucidate the spatial transcriptomic profile of adenomyosis lesions with matched eutopic endometrial subregions, with a focus on specific cell types and tissue compartments. The study was designed to identify differentially expressed genes, functionally enriched pathways, and cell-cell communication networks within adenomyosis lesions compared to matched endometrial subregions, and to nominate potential drug targets for future therapeutic development. We show that adenomyosis lesions exhibit a distinct transcriptome intermediate between matched endometrium and myometrium, with enhanced epithelial ciliation and altered biological pathways. We reveal that lesion transcriptome is more like the endometrial basalis than the functionalis, displaying increased olfactory signalling, altered oxidative phosphorylation and ATP synthesis, and an altered immune microenvironment suggestive of chronic inflammation. In silico drug screening identified potential compounds that can reverse the adenomyosis lesion transcriptome. These findings provide insights into the molecular landscape of adenomyosis lesions and pave the way for developing novel, lesion-specific treatments that could potentially spare the eutopic endometrium and offer alternatives to hysterectomy. Overall design: Full-thickness endometrial biopsies and matched myometrium containing adenomyosis lesions were prospectively collected from ten premenopausal women undergoing hysterectomy for pelvic pain and/or heavy menstrual bleeding. Paraffin-embedded human uterine biopsies underwent NanoString GeoMx digital spatial profiling. NanoString GeoMx DSP uses a pool of ~20,000 transcript-specific RNA probes, links each of the probes to a unique, probe specific barcode through a UV cleavable linker. The probe pool was hybridized to target mRNAs on tissue slides (overnight), and slides were then stained using fluorophore-labelled antibodies (“morphology markers”; here panCK, CD45, CD31) and imaged on the GeoMx profiler platform. A total of 6 regions of interest (ROIs) were selected per tissue sample (luminal epithelium, endometrial functionalis, endometrial basalis, adenomyosis lesions, and two myometrial regions), and UV light was applied to release and collect the DSP barcodes from selected ROIs. Photocleaved oligonucleotides from ROI aspirates, containing target-specific readout tags and unique molecular identifiers (UMIs), were used for library preparation. Libraries were quality-checked using a Bioanalyzer system (Agilent) and quantified with a Qubit high-sensitivity DNA kit. Sequencing was conducted on an Illumina NovaSeq system using NovaSeq S1 flow cells with paired-end 2 x 50 bp reads, generating approximately 1300 million clusters per cell. Digital count conversion (DCC) files were produced for each target probe.