The human body is not sterile. Beyond the trillions of bacteria that make up the microbiome, scientists now recognize a vast community of viruses, collectively called the virome, that inhabits tissues, mucosal surfaces and the bloodstream. Most of these viruses, particularly the herpesviruses, establish lifelong latent infections after the initial exposure and are normally held in check by the immune system. What has remained unclear is whether these resident viruses behave differently in people experiencing acute disease, and in particular whether their reactivation plays any role in stroke, a condition long known to involve inflammation but rarely studied through the lens of virology.

To interrogate this question, the team, led by Michael D. Clarke and senior author Glen C. Jickling, performed RNA sequencing on blood samples from 37 patients who had suffered an ischemic stroke, the type of stroke caused by a blocked blood vessel in the brain, and 32 age- and sex-matched controls with no stroke history. RNA sequencing reads the full complement of genetic transcripts in a sample, both human and viral. The researchers aligned the millions of sequencing reads against the human reference genome first to strip out host sequences, and then against a comprehensive curated database of human virus genomes to detect and quantify any viral transcripts present. Applying strict quality filters, requiring that a viral species be found in more than 3 percent of samples and with raw read counts above a threshold of 2, the team identified transcripts from six distinct viral families.

The list of detected viruses reads like a catalog of humanity’s most common passengers: several members of the human herpesvirus family, adenoviruses, papillomaviruses, human pegivirus, respiratory syncytial virus, and human endogenous retrovirus K, an ancient retroviral fossil embedded in our own genome that is occasionally transcribed. But the most striking finding concerned the herpesviruses. When the researchers pooled the read counts from all human herpesviruses together, they found that stroke patients carried 2.13-fold more herpesvirus transcripts than controls. In other words, these viruses were not merely present; they were more transcriptionally active, actively manufacturing viral RNA in the hours and days surrounding the stroke.

Just as notable was the pattern of coinfection. Many people carry a single herpesvirus, such as herpes simplex virus, cytomegalovirus, or Epstein-Barr virus, but harboring several simultaneously is less common. The study found that stroke patients had, on average, 1.23 times more distinct herpesvirus species detectable in their blood than controls. This simultaneous reactivation of multiple herpesviruses suggests a coordinated loss of immune surveillance rather than a chance flare-up of one virus, hinting that the immune system of stroke patients may be broadly distracted or dysregulated at the moment of the cerebrovascular event.

The analysis then drilled down to the level of individual viral genes, and two names emerged from the noise. The first was UL95, a gene from cytomegalovirus, a herpesvirus that infects the majority of adults worldwide and typically causes no symptoms in healthy hosts. The second was EBNA2, a gene from Epstein-Barr virus best known for its role in driving B-cell proliferation and its association with multiple sclerosis and several cancers. Transcripts from both of these genes were significantly increased in the blood of stroke patients compared with controls. EBNA2 is particularly interesting to immunologists because it acts as a master regulator of latency, rewiring the host cell’s transcriptional machinery and orchestrating immune evasion. Its elevated expression in stroke blood raises the possibility that EBV-infected cells are not passive bystanders during a stroke but are actively modulating the host response.

The most provocative part of the study came when the researchers examined what was happening to human gene expression in the same samples. They found that a set of host genes already implicated in stroke biology, including APOE, the famous Alzheimer’s and cardiovascular risk gene; C3, a central component of the complement cascade that bridges innate immunity and inflammation; PDGF, a growth factor involved in vascular repair; and CXCL2, a chemokine that recruits immune cells to sites of injury, were differentially expressed specifically in stroke samples that contained high counts of UL95 or EBNA2 transcripts. This correlation between active viral gene expression and altered host immune signaling provides the first tentative mechanistic thread connecting the blood virome to the inflammatory response after stroke. It suggests, though does not prove, that reactivating herpesviruses may help shape how the immune system responds to a cerebrovascular event, and perhaps even influence how that event unfolds.

The idea that infections can precipitate vascular events is not new. Epidemiological studies have repeatedly observed that acute infections, including influenza and bacterial pneumonia, are followed in the weeks afterward by transient spikes in heart attack and stroke risk. Vaccination against influenza has been associated with reduced stroke incidence in some studies, and chronic infections have been proposed as contributors to atherosclerosis itself. What has been missing is a direct, unbiased look at which viruses are actually active in stroke patients’ blood, and that is precisely what metagenomic RNA sequencing now makes possible. Rather than testing for a handful of preselected pathogens with antibody assays or targeted PCR, this approach captures everything with viral RNA in the sample, revealing the virome’s composition without preconceptions.

The technical achievement should not be understated. Viral transcripts are vanishingly rare in a blood transcriptome dominated by human hemoglobin genes, immune cell transcripts and ribosomal RNA. Detecting them requires deep sequencing, careful bioinformatic alignment against large viral reference databases, and conservative filtering to eliminate false positives from sequencing artifacts or contaminants. The authors’ use of prevalence and count thresholds reflects this caution, ensuring that only reproducibly detected viral sequences were analyzed. The linearization of viral genomes for read mapping, described in the study’s supplementary methods, further addresses the technical difficulty of aligning short sequencing reads against circular or complex viral genomes.

The researchers are careful about interpretation, and so should readers be. This is an observational study: it cannot determine whether herpesvirus reactivation contributes to causing strokes, whether the stress of an acute stroke somehow wakes dormant viruses from their latency, or whether both phenomena are driven by a shared upstream factor such as systemic inflammation or an immune-triggering event. All three scenarios remain biologically plausible. Herpesviruses are exquisitely sensitive to physiological stress hormones like cortisol, which can suppress cellular immunity and permit reactivation, so a stroke itself, or the conditions leading up to it, might plausibly trigger the viral activity observed. Conversely, a viral flare could promote a prothrombotic, proinflammatory state that raises the risk of a clot forming in a cerebral artery, a mechanism that has been proposed for other infections.

Future work will need longitudinal designs, following people over time to see whether rising herpesvirus transcription precedes stroke onset, and larger cohorts to confirm these findings across diverse populations. The Canadian and American research teams behind the study, funded by the Canadian Institutes of Health Research, the NIH, the Heart and Stroke Foundation of Canada and the University Hospital Foundation, suggest that such work could eventually open new therapeutic avenues. If viral reactivation turns out to modulate stroke risk or outcome, antiviral drugs or vaccination strategies might one day join the arsenal of stroke prevention tools. For now, the study stands as a vivid reminder that stroke is not purely a disease of blood vessels and brain tissue, but an event that unfolds across the entire ecological landscape of the human body, including the ancient viruses we have carried within us for a lifetime.

The findings also add to a growing literature implicating Epstein-Barr virus in neurological disease, following recent large-scale evidence linking EBV infection to multiple sclerosis. Whether the virus plays any comparable role in cerebrovascular disease will be a question for the next generation of studies. What this work establishes firmly is that the human blood virome is a real and measurable entity, that it shifts detectably in acute illness, and that ignoring it may mean missing a layer of biology that matters for stroke.