The study combined deCODE cis-pQTL summary data with the FinnGen R12 M13_INTERVERTEB endpoint, labelled “other intervertebral disc disorders.” After harmonization, 5,820 usable SNP rows supported screening across 1,723 genes. Investigators calculated single-SNP Wald ratios, inverse-variance weighted random-effects estimates when multiple variants were available, weighted-median estimates, MR-Egger slopes, heterogeneity statistics, Egger intercepts and leave-one-out estimates. Benjamini–Hochberg correction was applied to the 1,723 primary gene-level tests. The four multi-SNP signals that passed this threshold were MST1, with an odds ratio of 1.057, 95% confidence interval 1.038–1.077 and q = 4.15 × 10⁻⁶; DLK1, with an odds ratio of 0.933, 95% CI 0.903–0.964 and q = 0.016; ITIH1, with an odds ratio of 0.926, 95% CI 0.893–0.961 and q = 0.016; and SAA1, with an odds ratio of 0.927, 95% CI 0.892–0.962 and q = 0.023. These estimates describe associations between genetically proxied protein abundance and the recorded disc-disorder outcome, not direct effects on disc tissue.

Directional agreement was seen across several analytical methods for the four correction-positive genes: the weighted-median and MR-Egger estimates pointed in the same direction as inverse-variance weighting. Leave-one-out analyses also preserved the sign of each estimate when individual variants were omitted, and MST1 and SAA1 remained nominally significant after every displayed omission. Those checks provide internal consistency but cannot replace validation of instrument independence, particularly because the supplied pQTL table did not include a reference linkage-disequilibrium panel or an auditable record of variant clumping. The authors consequently described the results as local MR screening signals. TP53I3 and BPI were also examined because they were directionally concordant and present in the single-cell matrix, but their q values of 0.13 and 0.17 placed them only in an exploratory tier.

Human nucleus pulposus tissue supplied a cell-level view of the genetic candidates. The researchers analyzed GEO dataset GSE153066, which contains eight relatively normal libraries and eight libraries from degenerated discs. Quality control filtered the original 85,205 barcodes to 70,622 cells, retaining cells with more than 200 and fewer than 5,000 detected genes and less than 20% mitochondrial counts. Marker-informed annotation resolved eight broad populations: B cells, erythroid-like cells, fibroblasts, macrophages, natural killer cells, nucleus-pulposus chondrocyte-like cells, T cells and a deliberately unresolved stress population. Low-specificity clusters were not assigned forced identities, and the cell maps were treated as descriptive because individual cells from the same donor or library are not independent patient-level replicates.

Condition comparisons were instead anchored to the 16 libraries, with eight controls and eight degenerated samples treated as unpaired observations. Counts were summed within each library and transformed into log2 counts per million plus one for pseudobulk analysis. This design exposed substantial between-library heterogeneity and prevented the much larger number of cells from being mistaken for independent clinical evidence. The control group had a median age of 34.5 years, compared with 43 years in the degenerated group, but no age-adjusted or paired estimate was substituted for the unpaired screen. The dataset is cross-sectional, so its cell states cannot establish trajectories of degeneration or reveal whether a particular population caused a change over time.

Macrophages supplied the largest positive nominal difference in cell composition between degenerated and control libraries, amounting to 20.58 percentage points, with P = 0.031 and q = 0.250. The result did not survive correction within the composition panel and was therefore treated as exploratory rather than as an FDR-supported shift. Among 12 displayed follow-up genes, BPI had the largest positive library-level expression effect, with log2 fold change 2.932, P = 0.035 and q = 0.27. GCLM was also nominally positive, with log2 fold change 0.812, P = 0.044 and q = 0.27, but neither estimate met the within-panel false-discovery threshold. These findings indicate that some molecular differences may accompany degeneration while emphasizing that sample variability and multiple testing limit the strength of the conclusions.

A prespecified six-gene ferroptosis-related pQTL panel was analyzed separately from the four correction-positive multi-SNP genes. It included TFRC, FTL, AKR1C3, GCLM, HMOX1 and HSPB1, with a broader marker program that added GPX4, SLC7A11, ACSL4, FTH1, NFE2L2 and PTGS2 for expression context. TFRC was the only member supported in the genetic screen: a single-SNP analysis produced an odds ratio of 1.471, 95% CI 1.212–1.785, P = 9.42 × 10⁻⁵ and q = 0.023. Because this result rested on one SNP rather than a multi-SNP instrument, it supports follow-up but does not demonstrate that circulating transferrin receptor 1 is a causal driver of intervertebral disc disorders. The other panel members entered the tissue analysis as expression markers, independently of genetic significance.

The researchers also sought experimental support for selected findings in a human monocytic cell line. They differentiated THP-1 cells with 100 nM phorbol 12-myristate 13-acetate for 24 hours, allowed a 24-hour recovery, and then exposed the macrophage-like cells to 10 ng/mL each of interleukin-1 beta and tumor necrosis factor alpha for 24 hours. Relative to controls, the inflammatory condition produced higher MST1 expression, measured at 1.34 ± 0.07, higher BPI at 3.46 ± 0.28 and higher GCLM at 1.70 ± 0.12, while SAA1 expression was lower at 0.31 ± 0.03. Each comparison used four independent biological replicates per group and produced a two-sided Welch P value below 0.001. The directions were consistent with the corresponding library-level single-cell results, providing cross-platform concordance for four selected genes.

That laboratory agreement nevertheless has clear boundaries. THP-1 cells are a macrophage-like in vitro model, not primary nucleus-pulposus cells, and the PMA differentiation and recovery conditions can shape inflammatory responsiveness. The experiment therefore supports the plausibility of the measured expression directions in this selected system, not a mechanism of disc degeneration or a clinical effect in patients. Replication would require primary or disc-relevant models, prespecified stimulation conditions, independent biological replication and fuller qPCR reporting, including raw Ct values, primer details and verified amplification efficiency. The genetic analysis likewise needs independent instruments with documented linkage disequilibrium, broader outcome phenotypes, ancestry-diverse data and separate single-cell cohorts. Taken together, the study offers a layered candidate map linking circulating protein signals, cellular context and an inflammatory cell model, but it does not yet establish causal pathways, therapeutic targets or clinical interventions.

Subject of Research: Genetic and single-cell evidence for protein signals and ferroptosis-related states in human intervertebral disc degeneration

Article Title: Genetically supported protein signals and ferroptosis-related cell states in human intervertebral disc degeneration: local pQTL–FinnGen screening integrated with nucleus pulposus single-cell RNA sequencing

Article References: Tang, G., Wu, X., Lin, Y., Tang, Z., & Liang, H. (2026). Genetically supported protein signals and ferroptosis-related cell states in human intervertebral disc degeneration: local pQTL–FinnGen screening integrated with nucleus pulposus single-cell RNA sequencing. Frontiers in Cell and Developmental Biology, 14, Article 1944218. https://doi.org/10.3389/fcell.2026.1944218

Image Credits: AI Generated

DOI: 10.3389/fcell.2026.1944218

Keywords: intervertebral disc degeneration, Mendelian randomization, protein quantitative trait loci, nucleus pulposus, single-cell RNA sequencing, ferroptosis, TFRC, MST1, SAA1, THP-1 cells, FinnGen, Genetically

The study, published in the Journal of Translational Medicine, comes from a group led by Jingyuan Liu and Yang Liu at the Department of Radiotherapy of the Affiliated Cancer Hospital of Zhengzhou University and Henan Cancer Hospital, with collaborators at the same institution and at Zhengzhou University’s First Affiliated Hospital. Their central finding is deceptively straightforward. In mouse models of non-small cell lung cancer liver metastasis, a prophylactic dose of just 2 Gray, delivered to the liver before metastatic challenge, significantly suppressed metastatic colonization. Among the doses tested, 2 Gy showed the strongest antimetastatic effect, a dose low enough that it causes minimal tissue damage yet high enough to fundamentally alter how the liver communicates with the immune system.

What makes this result remarkable is the mechanism behind it. Radiation therapy is conventionally understood as a cytotoxic tool, a way of killing tumor cells with focused beams of high-energy photons. But over the past two decades, radiobiologists have come to appreciate that radiation does far more than damage DNA. At low doses, it can reprogram the behavior of cells in the irradiated field, altering the signals they send to immune cells. The Zhengzhou team set out to test whether such immune modulation could be harnessed prophylactically, not to destroy tumor cells directly, but to make the liver a hostile landing ground for them.

To answer this question, the researchers combined an unusually broad set of tools. They used bulk RNA sequencing to survey gene expression across irradiated liver tissue, single-cell RNA sequencing to resolve which individual cell types were responding, and in vitro functional assays to test the signaling pathways they uncovered. Finally, they performed in vivo neutralization experiments, blocking a key molecule with antibodies in living animals to confirm that the mechanism they had identified was genuinely responsible for the antimetastatic effect. This layered approach allowed them to trace a complete signaling chain from hepatocytes, the main functional cells of the liver, all the way to the recruitment of cytotoxic immune cells.

The chain begins with hepatocytes. When the liver received its 2 Gy dose, the team found that hepatocytes increased their production of serum amyloid A1, or SAA1, a protein more commonly associated with inflammatory responses. SAA1 then acted on a specific population of immune cells residing in or recruited to the liver: inflammatory monocytes. These monocytes, which in the context of an unprepared pre-metastatic niche can be co-opted by tumors to suppress immunity and scaffold metastatic growth, were instead pushed in a fundamentally different direction by the irradiation-induced signal.

At the molecular level, SAA1 engaged Toll-like receptor 2, a pattern-recognition receptor best known for detecting microbial components, on the surface of the monocytes. Engagement of TLR2 activated a transcriptional program driven by the NF-κB pathway, specifically through its RELA subunit, a master regulator of inflammatory gene expression. The consequence of this activation was the upregulation of SLC7A11, a solute carrier protein that imports cystine into cells and is a central component of cellular antioxidant defenses. With SLC7A11 activity elevated, the monocytes achieved improved redox homeostasis, a better-balanced internal oxidative state that allowed them to sustain a new functional identity.

That identity was chemokine-secretory. The reprogrammed monocytes became robust producers of CXCL10, a chemokine with a well-established role in attracting cytotoxic immune cells. In the irradiated liver, CXCL10 served as a beacon, drawing CD8-positive T cells, the principal killer lymphocytes of the adaptive immune system, into the tissue that tumor cells would soon attempt to colonize. When circulating lung cancer cells arrived, they encountered a liver patrolled by an alert cytotoxic army rather than one paved with immunosuppressive stromal support. The pre-metastatic niche had been flipped from pro-metastatic to anti-metastatic.

The team did not stop at correlation. To prove that CXCL10 was functionally required for the protective effect, they neutralized the chemokine in irradiated animals. The result was decisive: blocking CXCL10 reduced the recruitment of CD8-positive T cells to the liver and attenuated, though did not necessarily abolish, the antimetastatic benefit of low-dose irradiation. This loss-of-function experiment anchors the entire mechanistic model. Without the chemokine signal, the reprogrammed monocytes lose their ability to summon cytotoxic T cells, and the liver reverts toward vulnerability. SLC7A11, meanwhile, was shown to be required to sustain the chemokine-secretory phenotype itself, placing the antioxidant transporter upstream of CXCL10 production in the causal chain.

The translational implications are considerable. Liver metastasis from lung cancer is a major driver of mortality, and current strategies for preventing it are limited. Immune checkpoint blockade has transformed outcomes for some patients with non-small cell lung cancer, but hepatic metastases are notoriously resistant to immunotherapy, in part because the liver’s immune environment is inherently tolerogenic. A prophylactic intervention that remodels the hepatic immune landscape before tumor cells arrive could, in principle, complement existing approaches. The dose involved, a single 2 Gy fraction, is far below the levels used for ablative radiotherapy and is comparable to fractions routinely delivered in palliative settings, suggesting that toxicity would be manageable, though this remains to be demonstrated in patients.

Caution is warranted. The findings come from mouse models, and the authors note that the published version is an early-release, peer-reviewed accepted manuscript subject to further editorial processing. The retrospective clinical component of the study was approved by the Ethics Committee of the Affiliated Cancer Hospital of Zhengzhou University and Henan Cancer Hospital, and the animal work followed ARRIVE guidelines, but no prospective human trial data yet exist. Irradiating a healthy liver prophylactically raises legitimate questions that only clinical studies can answer, including optimal timing, patient selection, and long-term hepatic safety. Still, the study opens an intriguing frontier in what the authors describe as low-dose immune radiobiology: the idea that radiation, at doses far too low to kill tumor cells, can be deployed as an immune sculpting tool. If the hepatocyte-to-monocyte-to-CXCL10 axis identified here holds up in humans, a single low-dose treatment could one day help keep the liver off the map of metastatic destinations for lung cancer patients.

Subject of Research: Prophylactic low-dose liver irradiation reprogramming the pre-metastatic immune niche to suppress non-small cell lung cancer liver metastasis

Article Title: Prophylactic low-dose liver irradiation reprograms the pre-metastatic immune niche and suppresses NSCLC liver metastasis

Article References: Liu, J., Zhang, T., Zhu, X., Lan, Z., Li, L., Yang, Y., Wang, Q., Sun, Z., & Liu, Y. (2026). Prophylactic low-dose liver irradiation reprograms the pre-metastatic immune niche and suppresses NSCLC liver metastasis. Journal of Translational Medicine. https://doi.org/10.1186/s12967-026-08883-8

Image Credits: AI Generated

DOI: 10.1186/s12967-026-08883-8

Keywords: low-dose irradiation, pre-metastatic niche, liver metastasis, NSCLC, SAA1, TLR2, NF-kappaB, SLC7A11, CXCL10, inflammatory monocytes, CD8 T cells, radiobiology

The research team, led by Qianqian Yin and Yaru Yang of Xuzhou Central Hospital affiliated with Southeast University, together with colleagues including corresponding authors Xiaoyan Liu and Jianhua Zheng, recruited twenty patients diagnosed with PMOS and ten control women. Follicular fluid, the nutrient-rich liquid surrounding the oocyte inside each ovarian follicle, was collected during procedures related to in vitro fertilization and embryo transfer. Because this fluid is in direct contact with the developing egg and the surrounding cumulus cells, it offers a uniquely intimate window into the microenvironment in which fertilization competence is established. Rather than studying blood, where systemic changes can be diluted and confounded, the researchers chose to interrogate the immediate biochemical neighborhood of the egg itself.

The technological centerpiece of the study was four-dimensional label-free quantitative proteomics, an approach that has rapidly become one of the most powerful tools in modern biology. Traditional proteomic methods separate peptides along a single dimension of mass spectrometric analysis, but four-dimensional platforms add ion mobility separation and retention time alignment, dramatically increasing the depth, resolution, and accuracy of protein identification. Label-free quantification, meanwhile, avoids the chemical labeling steps that can introduce bias and expense, instead relying on the raw intensity of detected peptide signals to estimate relative protein abundance across samples. This combination allowed the team to comprehensively profile the whole proteome of follicular fluid and to detect subtle but meaningful differences between patient groups that earlier methods might have missed.

When the data were analyzed with strict statistical thresholds requiring a fold change greater than 1.5 and a p-value below 0.05, thirty-three proteins emerged as differentially expressed between the PMOS and control groups. Twenty-four of these proteins were elevated in the patients, while nine were reduced. That modest number may sound unremarkable, but each of these molecules represents a potential thread in the tangled web of PMOS pathogenesis, and the patterns revealed by subsequent bioinformatic analysis were strikingly coherent rather than random.

Functional enrichment analysis using the Gene Ontology framework showed that the altered proteins clustered around several biological themes. The first was immunoglobulin production and the humoral immune response, indicating that antibody-mediated immune activity is heightened within the follicular environment of PMOS patients. The second was leukocyte migration, suggesting that immune cells are being actively recruited into the ovarian follicle. The third and perhaps most surprising theme was cellular energy metabolism, with proteins involved in glycolysis and pyruvate metabolism prominently represented. These findings imply that the follicle in PMOS is not merely a passive site of hormonal dysfunction but an active immunometabolic battlefield where energy production and immune defense are simultaneously rewired.

Pathway-level analysis using the Kyoto Encyclopedia of Genes and Genomes reinforced this picture. The differentially expressed proteins were predominantly enriched in glycolysis and gluconeogenesis, pyruvate metabolism, the hypoxia-inducible factor-1 signaling pathway, and extracellular matrix-receptor interactions. The involvement of HIF-1 signaling is particularly intriguing. This pathway is the cell’s master response to low oxygen, and its activation can drive cells to shift their energy production from efficient mitochondrial respiration toward rapid but wasteful glycolysis, a phenomenon famously exploited by cancer cells. Its enrichment in PMOS follicular fluid raises the possibility that the ovarian follicle in these patients exists in a state of relative hypoxia or metabolic stress, forcing the egg and its supporting cells to operate under conditions that may compromise developmental potential.

Protein-protein interaction network analysis added another layer of insight, suggesting a synergistic relationship between metabolic reprogramming and immune dysregulation in PMOS. In other words, the altered energy metabolism and the disturbed immune signaling do not appear to be independent phenomena but are likely intertwined, each feeding into and amplifying the other. This kind of systems-level view is exactly what proteomics is designed to deliver, and it points toward a conceptual model in which PMOS arises from a self-reinforcing loop of inflammatory and metabolic disturbance within the ovary, rather than from a single defective gene or hormone.

To ensure that the mass spectrometry findings were not artifacts, the team turned to enzyme-linked immunosorbent assay, a targeted and highly specific method for measuring individual proteins. Three candidates were selected for validation: L-lactate dehydrogenase A chain, known as LDHA, a key enzyme that drives the final step of glycolysis and converts pyruvate into lactate; protein S100-A9, or S100A9, an inflammatory mediator released by neutrophils and monocytes that acts as a danger signal to the immune system; and serum amyloid A1, or SAA1, an acute-phase protein classically associated with inflammation and tissue remodeling. All three were confirmed to be significantly higher in the follicular fluid of PMOS patients than in controls, lending robust analytical credibility to the discovery set and highlighting a concrete biochemical axis linking glycolytic flux, inflammation, and acute-phase response within the follicle.

The implications of this work are twofold. Clinically, the identified proteins could serve as diagnostic biomarkers, offering a way to identify or monitor PMOS through a simple assessment of follicular fluid, and potentially as therapeutic targets whose modulation might restore a healthier follicular environment. Scientifically, the study reframes PMOS as a disorder of immunometabolic reprogramming and extracellular matrix remodeling, providing a new perspective on pathogenesis that goes beyond the traditional focus on androgen excess and insulin resistance. The authors are careful to note the limitations of their work: the sample size was modest, with twenty patients and ten controls, and the findings now require validation in larger cohorts, ideally integrated with detailed clinical phenotypes and functional experiments that can test whether the candidate proteins play causal roles or are merely markers of the disease process.

Nevertheless, the study stands as a compelling example of how cutting-edge proteomic technology can illuminate diseases that have long resisted conventional investigation. By capturing a snapshot of the protein milieu surrounding the human egg, Yin, Yang, and their colleagues have transformed PMOS from a vaguely defined clinical syndrome into a condition with a measurable molecular fingerprint. If subsequent research confirms and extends these findings, the glycolytic enzymes, inflammatory proteins, and matrix components identified in follicular fluid could become the foundation for new diagnostic tests and, ultimately, new treatments aimed at protecting the delicate metabolic environment in which every human life begins.

Subject of Research: Proteomic profiling of follicular fluid in polyendocrine metabolic ovarian syndrome

Article Title: Identification of differentially expressed proteins in follicular fluid from patients with polyendocrine metabolic ovarian syndrome using 4D label-free proteomics

Article References: Yin, Q., Yang, Y., Yan, X., Cao, Y., Liu, X., & Zheng, J. (2026). Identification of differentially expressed proteins in follicular fluid from patients with polyendocrine metabolic ovarian syndrome using 4D label-free proteomics. Journal of Ovarian Research. https://doi.org/10.1186/s13048-026-02271-y

Image Credits: AI Generated

DOI: 10.1186/s13048-026-02271-y

Keywords: polyendocrine metabolic ovarian syndrome, follicular fluid, proteomics, 4D label-free quantification, glycolysis, HIF-1 signaling, immune dysregulation, LDHA, S100A9, SAA1, biomarkers, fertility