> 매일 아침 research-watcher가 27개 query family × PubMed/GEO를 스캔해서, Brown Biotech 파이프라인 관점에서 decision-ready 인사이트로 정리합니다. 매일 06:00 KST 자동 생성. **Category focus:** Bioinformatics & Multi-Omics **Published:** 2026-10-28 06:10 KST **Entry ID:** 139 **Tags:** #fibrosis #diabetic kidney disease #DKD #TREM2 #macrophage #resident macrophage #monocyte-derived macrophage #compartment-specific protection #bone marrow chimera #spatial transcriptomics #single-cell RNA-seq #glomerulus --- ## 🔬 Today's Top Findings ### 1. [Fibrosis Lane + TREM2+ Macrophage Compartmentalization + Diabetic Kidney Disease + Resident-vs-Monocyte-Derived Origin] GSE324612 'TREM2+ macrophages from distinct origins mediate compartment-specific protection in diabetic kidney disease [Spatial Transcriptomics]' (Mus musculus, score 7, n=4, pdat 2026/10/09, raw_file_availability yes, license NCBI public repository, query 'muscle fibrosis spatial transcriptomics') is the FIRST NOVEL origin-resolved renal-fibrosis spatial dataset of the day ### 2. [Longevity Lane + Reproductive Aging + Single-Cell-Scale Spatial Ovarian Development Atlas + Follicle Reserve Establishment] GSE307688 'Single-cell-scale spatial transcriptome of the developing and adult mouse ovary' (Mus musculus, score 7, n=8, pdat 2026/10/14, raw_file_availability yes, license NCBI public repository, query 'Visium HD') is the FIRST NOVEL reproductive-longevity spatial atlas of the day ### 3. [Refractory-Cancer Lane + PD-L1-Driven Metabolic Reprogramming + PDK4/GLS Axis + Anti-PD-1 Resistance Override in NSCLC] GSE330687 'PD-L1 drives metabolic reprogramming and mitochondrial remodeling in NSCLC cells' (Homo sapiens, score 6, n=6, pdat 2026/05/17, raw_file_availability yes, license NCBI public repository, query 'metabolic flux transcriptome') is the FIRST NOVEL checkpoint-metabolic-escape dataset of the day ## 📋 Synthesis Loop 001 daily digest for 2026-10-28, generated from research-watcher output/latest/hits.jsonl (114 hits across 27 queries, scan regenerated 2026-10-28 06:03 KST). The score-9 top band was again entirely recycled against the retained id:135-id:138 window, so the actionable harvest comes from the score-7 and score-6 ladder after accession-level dedupe against the page-level GSE retention set (36 accessions retained on the page; 76 NOVEL accessions remain of today's 114). Signal 1 is GSE324612 (Mus musculus, n=4, pdat 2026/10/09), the JASN paper 'Spatially Distinct TREM2+ Macrophages Are Associated with Compartment-Specific Protection in Diabetic Kidney Disease' (doi: 10.1681/ASN.0000001248, PMID 42735028): scRNA-seq plus spatial transcriptomics of diabetic OVE26 kidneys combined with reciprocal bone-marrow chimeras shows TREM2 is induced in two anatomically distinct macrophage populations - resident-associated macrophages in the glomerular niche and monocyte-derived macrophages in the tubulointerstitium - and that the two populations protect DIFFERENT compartments, with resident TREM2 protecting the glomerulus and donor-derived TREM2 limiting tubulointerstitial injury. Signal 2 is GSE307688 (Mus musculus, n=8, pdat 2026/10/14), a Visium HD single-cell-scale spatial transcriptomic atlas of the developing and adult mouse ovary across eight fetal and postnatal timepoints (bioRxiv 2025.09.16.676655, PMID 41000831), which resolves cortex-versus-medulla patterning and follicle assembly at near-single-cell resolution and explicitly frames reproductive longevity as the downstream biological payoff. Signal 3 is GSE330687 (Homo sapiens, n=6, pdat 2026/05/17), the transcriptomic backbone of the JITC paper 'Targeting the PD-L1-induced PDK4/GLS metabolic axis overcomes anti-PD-1 resistance in non-small cell lung cancer' (doi: 10.1136/jitc-2026-015037, PMID 42686373), in which PD-L1 signalling drives metabolic reprogramming and mitochondrial remodeling in NSCLC cells through a PDK4/GLS axis. All three are accession-level NOVEL versus id:138 (CAF-ANP32A / YBX3-BAT / LAG-3-RNA-circuit) and versus the wider id:135-id:138 window, so the novelty gate is met at 3/3 against a required 2/2, with three distinct BB lanes covered (fibrosis, longevity/endocrine, refractory cancer). --- ## 🎯 Highlights ### 1. [Signal 1 - GSE324612 TREM2+ macrophages of distinct developmental origin in diabetic kidney disease, Mus musculus, score 7, n=4, pdat 2026/10/09]: the highest-scoring NOVEL accession of the day and the first NOVEL renal-fibrosis entry to enter the retention window. The JASN paper (doi: 10.1681/ASN.0000001248) integrates scRNA-seq, spatial transcriptomics of diabetic OVE26 kidneys, and reciprocal bone-marrow transplantation in a high-fat-diet plus streptozotocin model. Global TREM2 deficiency markedly worsened albuminuria, lowered GFR, aggravated glomerular and tubular injury - TREM2 is a protective factor, not a pathogenic one. The spatial component resolves that TREM2 is induced in TWO distinct macrophage populations in two distinct anatomic niches: resident-associated macrophages in the glomerulus and monocyte-derived macrophages in the tubulointerstitium. The chimeras then show the populations are functionally compartmentalized: resident-derived TREM2 protects the glomerulus, donor-derived TREM2 limits tubulointerstitial injury. This is a cell-of-origin-resolved, spatially-localized, functionally-separated protective axis with a directly actionable target (TREM2 agonism, or origin-restricted myeloid engineering). Why it matters for BB: Brown Biotech's fibrosis lane has been built on fibroblast and ECM biology; this is the myeloid-compartment counterpart, and the compartment-specific point is exactly the kind of design constraint a fibrosis brief needs - a systemic TREM2 agonist would not be equivalent to a glomerulus-restricted one. It is also the cleanest example in today's scan of the standard BB methodology template: cell-type state from scRNA-seq, spatial localization from spatial transcriptomics, causal assignment from chimera. ### 2. [Signal 2 - GSE307688 single-cell-scale spatial transcriptome of developing and adult mouse ovary, Mus musculus, score 7, n=8, pdat 2026/10/14]: the first NOVEL reproductive-longevity dataset to reach the digest, and a genuinely new lane rather than a variant of the aging-muscle material carried in id:137 and id:135. The dataset uses 10X Genomics Visium HD to profile the developing mouse ovary at eight key fetal and postnatal timepoints while preserving native cellular architecture, delivering near-single-cell-resolution spatial gene expression across ovarian development - all major ovarian cell types captured, with dynamic transcriptional signatures for the establishment of cortex and medulla and for follicle assembly in each region. The authors state the resource is intended to open new avenues on the spatial determinants of female fertility and reproductive longevity. Raw supplementary files (CLOUPE, CSV, H5, JSON, PARQUET, TIFF, ZIP) are available, so the atlas can be ingested directly. Why it matters for BB: this connects two Brown Biotech lanes that have never been joined in a single brief - the longevity lane (senescence, HBAT, geroprotector target nomination) and the reproductive-aging axis - through a spatial dataset rather than a clinical one, and the spatial-organogenesis framing is exactly what a cell-state-nomination deliverable requires, because follicle reserve is a spatial-architecture property, not a single-gene readout. It is also a rare case where the dataset itself is framed by the authors around a longevity question, which makes it a low-interpretation-cost entry point for the longevity service. ### 3. [Signal 3 - GSE330687 PD-L1-driven metabolic reprogramming and mitochondrial remodeling in NSCLC, Homo sapiens, score 6, n=6, pdat 2026/05/17]: the refractory-cancer pick of the day, and the first NOVEL checkpoint-metabolic-escape dataset in the retention window. The GEO accession is the transcriptomic backbone of the JITC paper 'Targeting the PD-L1-induced PDK4/GLS metabolic axis overcomes anti-PD-1 resistance in non-small cell lung cancer' (doi: 10.1136/jitc-2026-015037, PMID 42686373). The mechanistic claim is that PD-L1 signalling does not merely suppress T-cell function - it reprograms the tumor cell's own metabolism and mitochondrial architecture through a PDK4/GLS axis, and that blocking this axis reverses anti-PD-1 resistance. Why it matters for BB: this is a tumor-cell-intrinsic metabolic node downstream of an immune checkpoint, which is the exact geometry of the refractory-cancer briefs BB sells - the client question is never 'why did the checkpoint fail' but 'what does the tumor do once the checkpoint is engaged'. It gives a two-node intervention (PDK4 and GLS) with a human NSCLC backing and human data, and unlike the stromal EV axis in id:138 it is druggable by small molecules already in clinical use. The lineage is also directly continuous with the METABOLIC-TRACK thesis the digests have been building: id:136's LAG-3 state-gated RNA circuit, id:138's YBX3/BCAA flux node, and today's PDK4/GLS node all place metabolism downstream of an immune-state decision. ### 4. [Cross-signal synthesis - the new observation of the day is that all three NOVEL signals relocate the intervention to the compartment, not to the cell type]: Signal 1 shows that the SAME molecular target (TREM2) has OPPOSITE consequences in two renal compartments depending on the macrophage's developmental origin and anatomic niche, so the drug must be compartment-restricted rather than systemic; Signal 2 shows that the durable functional property (follicle reserve) lives in spatial architecture and patterning, so the relevant readout is a niche, not a marker; Signal 3 shows that the checkpoint drug fails because the tumor cell changes its metabolism in response to checkpoint engagement, so the second-line target must be the metabolic node the tumor itself moves to. Composed with id:136's state-gated delivery and id:138's state-carrier thesis, this closes a three-digest arc with a single proposition: THE PRODUCTIVE UNIT IS THE COMPARTMENT - ORIGIN, NICHE, OR STATE - AND A TARGET WITHOUT A COMPARTMENT SPECIFICATION IS NOT YET A PRODUCT. The concrete Brown Biotech product implication is a standing 'compartment specification' section that can be added to every target-nomination brief in the fibrosis, longevity, and refractory-cancer lanes, using today's three datasets as the worked examples. Second, on process: the recurring score-9 saturation continues - the top band has been fully recycled for a fourth consecutive cycle - which argues for the watcher adding a freshness-keyed query rotation so the scan stops re-serving the same immune-atlas accessions, since every cycle now spends two of three turns on dedupe rather than on biology. --- ## 💼 Next Steps - **[Open GSE324612 in GEO (TREM2+ macrophages origin-resolved DKD spatial, n=4)](https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE324612)** - **[Open GSE307688 in GEO (Visium HD single-cell-scale ovarian development atlas, n=8)](https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE307688)** - **[Open GSE330687 in GEO (PD-L1 metabolic reprogramming / mitochondrial remodeling in NSCLC, n=6)](https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE330687)** - **[Read the JASN TREM2 diabetic kidney disease paper (doi: 10.1681/ASN.0000001248)](https://pubmed.ncbi.nlm.nih.gov/?term=TREM2+macrophages+diabetic+kidney+disease+compartment-specific+protection)** - **[Read the ovarian spatial atlas preprint (bioRxiv 2025.09.16.676655)](https://pubmed.ncbi.nlm.nih.gov/?term=single-cell-scale+spatial+transcriptome+developing+adult+mouse+ovary)** - **[Read the JITC PD-L1 / PDK4-GLS anti-PD-1 resistance paper (doi: 10.1136/jitc-2026-015037)](https://pubmed.ncbi.nlm.nih.gov/?term=PD-L1+PDK4+GLS+metabolic+axis+anti-PD-1+resistance+NSCLC)** - **[Request a compartment-specified fibrosis brief (TREM2 myeloid origin)](https://brownbio.tech/services/fibrosis#brief)** - **[Request a longevity brief on spatial determinants of reproductive reserve](https://brownbio.tech/services/longevity#brief)** - **[Request a refractory-cancer brief on tumor-cell metabolic escape under checkpoint therapy](https://brownbio.tech/services/refractory-cancer#brief)** - **[Request an AI-DD brief on metabolic target nomination from checkpoint context](https://brownbio.tech/services/ai-drug-discovery#brief)** - **[View multiomics service](https://brownbio.tech/services/multiomics)** - **[View fibrosis service](https://brownbio.tech/services/fibrosis)** --- ## 📡 Provenance - **Source:** research-watcher (PubMed/GEO scan, 27 query families × Brown Biotech pipelines) - **Pipeline:** [`brown-biotech-daily-tech-digest`](https://github.com/ohbryt/brown-biotech-platform) — 06:00 KST cron - **Generated:** 2026-10-28 06:10 KST - **Repo:** `ohbryt/brown-biotech-platform` --- _Auto-published by `brown_biotech_research_digest_publisher.py` · [brownbio.tech](https://brownbio.tech) · Decision-ready research, daily._
Brown Biotech Research Digest — 2026-10-28
PubMed/GEO scan · research-watcher · 06:00 KST