Yina Du, Minzhe Guo, Yixin Wu, Andrew Wagner, Anne Karina Perl, Kathryn A Wikenheiser-Brokamp, Jane J Yu, Nishant Gupta, Elizabeth Kopras, Vera P Krymskaya, Kseniya Obraztsova, Yan Tang, David J Kwiatkowski, Elizabeth P Henske, Francis X McCormack, Yan Xu
8 min
Lymphangioleiomyomatosis (LAM) is a rare, progressive disease characterized by the cystic destruction of the lung, primarily affecting women. Despite the identification of TSC1/TSC2 mutations as key drivers, translating molecular insights into effective treatments remains challenging. A major obstacle has been the accessibility of complex single-cell omics data, which typically requires specialized computational skills to process and interpret.
To overcome these barriers, the authors developed the LAM Cell Atlas (LCA), a centralized, web-based repository. The portal aggregates data from 11 single-cell RNA-seq, 4 single-nucleus RNA-seq, and 3 single-nucleus ATAC-seq datasets. These samples were obtained from the lungs, uterus, and kidneys of LAM patients across three major research institutions. The platform provides a user-friendly, code-free interface for researchers to explore the cellular composition, gene regulatory programs, and ligand-receptor interactions within LAM lesions.
The LCA is organized into five functional modules: "LAM Transcriptome," "LAM Epigenome," "LAM Connectome," "LAM Projects," and "LAM Images." By integrating these multi-omics layers, the tool allows users to identify cell-specific markers, visualize chromatin accessibility, and predict cell-cell communication networks. As a demonstration of its utility, the authors performed a comparative analysis across lung, uterus, and renal samples. This identified both universal biomarkers and organ-specific regulatory pathways, highlighting the potential for discovery of novel diagnostic targets and therapeutic candidates.
By democratizing access to complex multi-omics data, the LCA serves as a "living" resource for the research community. It facilitates the testing of new hypotheses without the need for extensive bioinformatics expertise. The inclusion of a comprehensive image gallery further links molecular findings with tissue morphology, providing a holistic view of the disease. The atlas is intended to be iteratively updated, serving as a cornerstone for future collaborative efforts to understand LAM and design organ-specific, targeted treatments.
Lymphangioleiomyomatosis (LAM) is a rare lung disease of women, causing cystic remodelling of the lung and progressive respiratory failure. The cellular composition, microenvironment and cellular interactions within the LAM lesion remain unclear. To facilitate data sharing and collaborative LAM research, we performed an integrative analysis of single-cell data compiled from lung, uterus and kidney of patients with LAM from three research centres and developed an LAM Cell Atlas (LCA) Web-Portal. The LCA offers a variety of interactive options for investigators to search, visualise and reanalyse comprehensive single-cell multiomics data sets to reveal dysregulated genetic programmes at transcriptomic, epigenomic and cell-cell connectome levels.
Alex: Walk me through the central idea without turning this into a software tour. What does bringing those measurements together let someone ask?
Sam: Think of linked views of the disease. Gene activity helps identify cell populations and their signatures. Chromatin accessibility, meaning how open different regions of packaged DNA are, adds clues about gene regulation. Predicted communication networks suggest which cell types might signal to each other.
Alex: Those sound complementary, but they are not interchangeable. Does a gene activity pattern itself establish how that gene is regulated?
Sam: No. The atlas offers different measurements and analyses, not a guarantee that one proves another. It contains RNA data from cells and isolated nuclei, plus DNA accessibility data from nuclei. The authors use these to investigate related questions at different molecular levels.
Alex: Give me one example where a researcher could follow a question across the affected organs.
Sam: The authors independently identified what they call LAM core cells in the different tissues. These are the disease-associated cell populations they compare. Their cross-organ analysis found both shared gene signatures and signatures unique to a tissue.
Alex: A shared signature could be useful, but what concrete example do they show?
Sam: They highlight a gene called PCP4 as a novel marker. It was selectively expressed in LAM core cells from lung and uterus. It was also expressed in a particular cell population from a kidney angiomyolipoma, a tumor included in their comparison.
Alex: If the same marker appears across those tissues, can we call it a diagnostic marker for LAM?
Sam: Not from this evidence alone. The authors describe potential universal biomarkers with possible diagnostic, prognostic, and predictive utility. But the cross-organ comparison is explicitly proof of concept. Showing a shared expression pattern is not the same as establishing clinical performance.
Alex: So it nominates a candidate. What does the DNA accessibility view add beyond that kind of expression search?
Sam: Their example concerns ACTA2, another gene examined in the atlas. Its promoter, a region involved in controlling gene activity, was more accessible in LAM core cells than in the comparison mesenchymal cells. Those are cells from the tissue’s supporting compartment. The comparison offers a clue to regulatory differences within a LAM lung sample.
Alex: And the communication map sounds particularly easy to overinterpret. What is actually measured there?
Sam: The communication links are predictions from ligand–receptor analysis. That pairs signals cells may send with receptors other cells may carry. The portal lets users select a signaling family and a cell type, then inspect the predicted connections and their direction.
Alex: That is a map of plausible conversations, not evidence that a message was received.
Sam: That is the boundary to keep. The paper calls them predicted signaling interactions. Alongside these molecular tools, the atlas includes tissue images and project examples, helping users connect molecular questions with disease morphology.
Alex: Let’s put scale behind the resource. How much material supports all this exploration?
Sam: The portal contains about ninety-eight thousand cells. That is a substantial collection for exploration, assembled across the participating centers. But the number of cells is not the number of independent tissue samples.
Alex: That distinction matters especially when comparing organs. How much tissue replication is there?
Sam: The Discussion describes thirteen tissue samples in total. Most came from lungs. It specifies eleven LAM lungs, alongside a single uterine sample and a single kidney tumor sample.
Alex: So the broad cell count should not make the cross-organ comparison sound broadly replicated. Each non-lung organ is represented by just one sample.
Sam: That is my main caution too, and it matches the authors’ limitation. They call the organ comparison preliminary because LAM is rare and fresh tissue is difficult to obtain. They say more organ-specific data and experimental validation are needed.
Alex: They also discuss organ-specific treatment. How far does this study take that idea?
Sam: It proposes a rationale, not a tested therapy. Shared pathways may point toward common disease mechanisms. Tissue-specific pathways may reflect different local environments. The authors propose that understanding those differences could help design targeted, organ-specific treatments with fewer harmful effects.
Alex: Then who should read the full document, and where should they begin?
Sam: LAM researchers looking for markers, regulatory hypotheses, or candidate cell interactions should read it. Start with the application examples and the lung–uterus–kidney comparison. Then read the Discussion’s limitations and the supplementary methods before treating any candidate as established.
Alex: And for researchers outside LAM, what is the point to carry away?
Sam: The resource changes who can ask questions of these scarce samples. It does not, by itself, settle those questions. An atlas can make rare disease data accessible without making preliminary biology conclusive.
Alex: Thanks for listening.