The research team, led by Liang Cao and Mengzhou Xue together with V. Wee Yong, took an unusually comprehensive route to this discovery. Rather than relying on any single line of evidence, they wove together large-scale human genetics, transcriptomics, proteomics, single-cell resolution analysis, and animal experimentation. Their starting point was a familiar frustration in stroke genetics: genome-wide association studies have identified numerous risk loci for intracerebral hemorrhage, but most of these variants sit in non-coding regions of the genome, leaving the actual causal genes and mechanisms shrouded in ambiguity.

To cut through that ambiguity, the investigators deployed transcriptome-wide and proteome-wide association studies, known as TWAS and PWAS, which predict how genetically determined differences in gene expression or protein abundance influence disease risk. They integrated summary statistics from four large hemorrhage GWAS datasets, including a European-ancestry meta-analysis drawing on the FinnGen consortium and UK Biobank, independent FinnGen replication rounds, and the BioBank Japan cohort for trans-ethnic validation. Bayesian colocalization analysis then tested whether the genetic signals for disease and for gene regulation shared the same causal variant, while summary-data-based Mendelian randomization with the HEIDI test helped distinguish genuine causal effects from linkage artifacts.

Several genes emerged from this gauntlet, including ICA1L, WDR12, BAIAP3, MLYCD, CSPG5, SUN2, HABP4, and BID, several of which had never before been linked to hemorrhagic stroke. But one candidate stood apart. In the FinnGen R11 cohort, genetically predicted abundance of the LAMC1 protein was significantly associated with lower hemorrhage risk in two independent brain proteome datasets, and MAGMA gene-level analysis validated the signal. Most tellingly, when the researchers intersected their results across European and East Asian ancestry cohorts using two independent prioritization methods, LAMC1 was the only gene shared between populations, a convergence that substantially strengthens the case for its biological relevance.

The causal evidence deepened from there. Colocalization analysis in two independent brain eQTL datasets, BrainMeta v2 and PsychENCODE, yielded posterior probabilities of 0.923 and 0.934 that the LAMC1 expression signal and the hemorrhage risk signal share a single causal variant, with the lead SNP rs12146099 anchoring both signals on chromosome 1. Similar colocalization appeared in blood-based datasets from GTEx v8 and eQTLGen. Mendelian randomization across all four datasets pointed consistently in the protective direction: each one-standard-deviation increase in genetically predicted LAMC1 expression was associated with an odds ratio for hemorrhage well below 1.0, ranging from 0.873 down to 0.630 depending on the dataset.

Perhaps the most technically impressive step was pinpointing which brain cells actually matter. Because bulk brain tissue masks the signals of its highly heterogeneous cell populations, the team turned to single-cell expression quantitative trait loci datasets and ran colocalization analysis in every major cell type. The result was unambiguous: the genetic signal regulating LAMC1 expression and the hemorrhage risk signal colocalized most strongly within astrocytes, with a posterior probability of 0.956 in the FinnGen R11 analysis. Other genes told different stories, with ICA1L and WDR12 signals localizing to neurons, but LAMC1’s story was one of astrocytes and blood vessels.

Mouse experiments then brought the human genetics to life. Using a collagenase-induced hemorrhage model, the researchers tracked LAMC1 protein over two weeks and found it plummeted after injury, reaching its lowest point on day three, even though mRNA levels stayed flat, hinting that the protein is degraded rather than transcriptionally silenced. Multiplex immunofluorescence revealed exactly where the loss occurred: LAMC1 colocalized with the astrocyte marker GFAP and the endothelial marker CD31, and its signal dropped significantly in both cell types after hemorrhage. Crucially, it showed essentially no overlap with microglia or neurons, indicating a cell-type-specific collapse rather than a generic stress response.

The therapeutic proof came from gene delivery. Three weeks before inducing hemorrhage, mice received adeno-associated virus vectors driving LAMC1 overexpression under a GFAP promoter in the striatum. When hemorrhage was later induced, the treated animals fared markedly better. H&E staining showed preserved striatal architecture and reduced lesion area, brain water content fell significantly in the ipsilateral cortex and basal ganglia, and behavioral tests confirmed functional benefit: mice overexpressing LAMC1 scored higher on the forelimb placing test and stayed on the rotarod significantly longer than controls on day three.

The mechanism traces back to the blood–brain barrier. In an Evans blue extravasation assay, dye that should have been confined to vessels leaked extensively into the injured hemisphere of control mice, but LAMC1-overexpressing animals showed significantly less leakage. Western blotting revealed why: hemorrhage drastically reduced the tight junction proteins ZO-1 and occludin, which seal the gaps between endothelial cells, and LAMC1 overexpression reversed that loss. Immunofluorescence showed the same story visually, with tight junction staining that was fragmented and discontinuous in controls but restored to continuous, linear vascular profiles in treated mice. LAMC1, a laminin gamma chain glycoprotein normally embedded in the basement membrane, appears to physically scaffold the barrier and prevent degradation of its molecular seals.

The translational implications are notable. Unlike neuronal risk genes such as ICA1L and WDR12, which act intracellularly, LAMC1 is an extracellular matrix protein secreted by astrocytes and endothelial cells, making it a more accessible drug target. The authors suggest strategies such as blocking matrix metalloproteinase-mediated degradation to stabilize endogenous LAMC1, or administering recombinant laminin mimetics during the acute phase, potentially targeted at patients genetically stratified as low LAMC1 expressers. Important questions remain, including which proteases drive LAMC1 degradation and how environmental factors like chronic hypertension interact with genetic risk. But the study delivers something hemorrhagic stroke research has rarely had: a genetically validated, mechanistically explained, experimentally confirmed protective target, and a demonstration that shoring up the neurovascular unit from the astrocyte side may be a viable way to blunt the deadliest stroke of all.

Subject of Research: Genetic and cell-type-specific mechanisms of protection against intracerebral hemorrhage mediated by astrocyte-derived LAMC1

Article Title: Astrocyte-derived LAMC1 protects against intracerebral hemorrhage: A novel genetic mechanism maintaining neurovascular integrity

Article References: Cao, L., Zhang, Y., Pi, W., Zhang, R., Zhang, Y., Yong, V., & Xue, M. (2026). Astrocyte-derived LAMC1 protects against intracerebral hemorrhage: A novel genetic mechanism maintaining neurovascular integrity. Journal of Advanced Research, 88, 867-884. https://doi.org/10.1016/j.jare.2026.01.028

Image Credits: AI Generated

DOI: 10.1016/j.jare.2026.01.028

Keywords: intracerebral hemorrhage, LAMC1, astrocytes, blood-brain barrier, Mendelian randomization, colocalization analysis, single-cell eQTL, tight junction proteins, laminin, stroke genetics, neuroinflammation, gene therapy