# Intestinal–vascular barrier (GVB): PV1, portal blood, and limits of clinical inference

- Canonical: https://nutrifit.health/knowledge/gut-vascular-barrier-gvb-pv1-liver-metastasis-leaks
- Language: en
- Published: 2026-10-09T04:11:13.823Z
- Updated: 2026-10-09T04:52:21.158Z
- Author: NutriFit Editorial

> How the epithelial and vascular boundaries differ, what PLVAP/PV1 marks, what mouse models show about Salmonella and MASH, and where human diagnostic and treatment claims stop.

A microbe does not reach the liver merely by crossing the epithelium. Between the intestinal lumen and portal circulation lie mucus, epithelial cells, the lamina propria with its immune cells, and the wall of a blood microvessel. The last boundary is called the gut–vascular barrier (GVB). Experimental work on it helped explain why bacterial access to subepithelial tissue and systemic dissemination are different events. But “open GVB” is not a diagnosis in its own right, and endothelial PV1 staining does not automatically measure the movement of live bacteria or an individual person’s risk of metastasis.

## Method and limitations

This is a narrative scientific review of the intestinal vascular barrier, its molecular mechanisms, and possible links with liver disease. It compares animal experiments and cellular and tissue measurements with limited human data. It is not a systematic review or a formal assessment of evidence quality.

A targeted check of primary publications on 9 October 2026 clarified the Spadoni findings: in the *Salmonella* model, the relevant pathway involved the SPI-2 secretion system and endothelial β-catenin signaling; the work did not establish a specific causal action of the AvrA effector. Key studies by Bertocchi and Mouries and their experimental designs were also checked. A publication previously cited to support a serum PLVAP threshold in cirrhosis concerned a bacterial protease, not GVB; therefore, no clinical threshold is reported. The remaining sources did not undergo a comprehensive independent audit.

PV1 in tissue is treated as a research marker of endothelial phenotype. Interventions in mice do not prove that the same treatment is effective in people, and tissue associations in humans do not establish causation. The studies reviewed do not establish a diagnostic threshold or treatment regimen for a standalone “GVB syndrome.”

## 1. Two barriers and two routes beyond the mucosa

### The epithelium is not the vascular endothelium

Secreted mucins, antimicrobial products, and the tightly joined single-layer epithelium are the first structures to encounter intestinal contents. Enterocytes are connected by epithelial tight junctions; disruption of those junctions changes the passage of water, ions, and molecules from the lumen into the lamina propria. Beneath it lies the endothelium of blood capillaries and venules. Its adherens and tight junctions, basement membrane, vesicular transport, and perivascular cells regulate the movement of material **from tissue into blood**. If the epithelium is damaged while the endothelium remains intact, those two steps can produce different results. Conversely, endothelial dysfunction alone does not show that a luminal antigen crossed the epithelium.

Blood from the intestine enters the portal vein through mesenteric veins and flows through liver sinusoids, where Kupffer cells and other cells remove some microbial products. Chylomicrons and some antigens can enter the central lacteals of villi and travel through mesenteric lymph nodes and then the thoracic duct. These routes are not completely separate: lymph eventually returns to systemic circulation, while portal blood passes through the liver. The claim that “every antigen is forced into lymph and never reaches the liver” would be wrong.

[![Four sequential compartments: lumen and mucus, epithelium, lamina propria, and vascular GVB; epithelial and endothelial flow require separate measurements.](https://api.nutrifit.health/api/v2/public/knowledge/files/images/nutrifit-scientific-20261009-gut-vascular-barrier-gvb-pv1-liver-metastasis-leaks-gvb-two-barriers-c186fdf513cb.png)](https://api.nutrifit.health/api/v2/public/knowledge/files/images/nutrifit-scientific-20261009-gut-vascular-barrier-gvb-pv1-liver-metastasis-leaks-gvb-two-barriers-c186fdf513cb.png)

Spadoni and colleagues demonstrated vascular selectivity and the possibility of its disruption in mice infected with *Salmonella enterica* serovar Typhimurium; human samples were also used to assess vascular changes associated with celiac disease and elevated aminotransferases. This does not show that every intestinal capillary in every healthy person is nonfenestrated or equally permeable. Vascular structure varies by intestinal segment, vessel type, age, and inflammatory state. The original claim of a rigid boundary—“4 kDa passes, everything larger never does”—describes a particular set of tracers and experimental conditions, not a universal law of intestinal exchange.

| Boundary | Structure | Experimental measurement | Conclusion that cannot be drawn automatically |
| :--- | :--- | :--- | :--- |
| Epithelial | Mucins, enterocytes, epithelial junctions | Tracer from lumen into mucosa, TEER, morphology | What specifically entered portal blood |
| Vascular GVB | Endothelium, junctions, transcellular route, basement membrane | Vascular tracer, CD31/CD34 and PV1 staining, microbes detected outside the intestine | A clinical diagnosis or cause of chronic disease |
| Lymphatic | Lacteals, valve-like/button junctions, lymph nodes | Lipid and cell labels, lymph collection | Absence of any portal signal |
| Hepatic | Sinusoids, Kupffer cells, hepatocytes | Portal/systemic samples, immune and tissue markers | Proof of a specific intestinal defect |

### Why vascular selectivity is different from “sealing”

The intestinal endothelium must protect the bloodstream from uncontrolled microbial access while allowing nutrients, hormones, water, metabolites, and immune signals to pass. A “barrier” therefore means regulated selectivity, not an absence of exchange. In experiments with fluorescent dextrans, passage of a molecule of a given mass depends on its hydrodynamic radius, shape, charge, and observation time. A particle with the same mass but a different conformation may behave differently. Vessel area, perfusion pressure, blood flow, and route of administration also affect the readout. Listing 4, 10, and 70 kDa thresholds without the methods does not define the absolute size of a pore.

It is useful to distinguish the paracellular route between endothelial cells from the transcellular route through them. The first involves VE-cadherin, claudins, and actin tension; the second involves endocytosis, vesicles, and diaphragmed structures. Inflammation can change both, but one tracer type cannot localize the route. Moreover, passage into a capillary wall does not guarantee detection in a systemic sample: liver and spleen cells rapidly remove some material. A quantitative conclusion requires agreement among local microscopy, tracer flux, and biological analysis of portal or peripheral blood.

### Cellular architecture of the endothelium

At adherens junctions, VE-cadherin (CDH5) forms calcium-dependent intercellular complexes and is linked to the actin cytoskeleton through p120-, β-, and α-catenins. Claudin-5 (CLDN5), occludin, JAM proteins, and ZO adaptors contribute to tight junctions. Their arrangement depends on the vascular segment; applying knowledge from the blood–brain barrier to the intestine requires direct testing. Disruption of the VE-cadherin complex can increase paracellular flux, but detecting the protein in a tissue section does not quantify permeability. Phosphorylation, endocytosis, localization, and junctional recovery dynamics matter as much as total expression.

The basement membrane contains type IV collagen, laminins, nidogen, and proteoglycans. Pericytes contact the endothelium and affect vessel stability, while lamina propria macrophages can take up particles before they approach the vessel wall. Enteric glia may modulate the epithelial and vascular environment, but the specific chain “glial GSNO → VE-cadherin in intestinal endothelium → fewer portal bacteria” requires intestinal vascular experiments; a study of the epithelial barrier cannot fill that gap. The same applies to pericyte-derived factors and MMP/TIMP: matrix proteolysis may occur during inflammation, but its role in the brain, a tumor, and normal intestine cannot be assumed to be the same.

A transcellular route also runs through endothelial cells. Caveolae containing caveolin-1, vesicular structures, and albumin transport have their own regulation; fenestrae and diaphragms are not equivalent to open gaps that admit bacteria. A lacteal has a different function and morphology, so GVB cannot be inferred from lacteal permeability, or vice versa. Blood-flow shear stress and the endothelial glycocalyx support vascular function; heparan sulfate, hyaluronan, and proteoglycans create a chemical and mechanical surface, but measured thickness and charge depend on the method and vascular bed.

### Spatial and age-related differences

A small-intestinal villus with a central lacteal, a crypt, and vessels in the colon differ in shape, immune-cell abundance, and microbial load. A venule used for leukocyte exit has a different set of functions from an exchange capillary. Tissue analyses should report the intestinal segment, depth, vessel type, and mucosal state; otherwise, different fields of view can create a false impression of “increased PV1.” Pediatric, adult, and inflamed tissues should not be combined into one reference range. The functional analogy between the GVB and the blood–brain barrier means that both involve regulated endothelium, not that they have identical CLDN5 expression, transport proteins, or degree of fenestration.

Inflammatory angiogenesis also deserves attention. New vessels in cancer, inflammatory bowel disease, or healing may express PV1 and have a different architecture from the original normal capillary wall. If their share increases, average tissue staining can rise without the pre-existing vessels necessarily “opening.” An initial tissue finding may therefore be a marker of remodeling, a causal contributor to permeability, or both; a time course and controlled intervention help distinguish these possibilities.

## 2. PV1/PLVAP: structure, staining, and causality

PV1, also called PLVAP, is a transmembrane glycoprotein associated with stomatal diaphragms of fenestrae, caveolae, and some transendothelial channels. Ultrastructural studies by Stan and colleagues established it as a useful structural marker of endothelial phenotype. In experimental GVB impairment, PV1-positive staining increases in relevant vessels. PLVAP is also expressed in other vascular beds, however, and in some contexts its diaphragms can **restrict** the passage of soluble proteins rather than simply increase it. A higher PV1 signal does not mean “a hole”: the location of the protein, the appearance of fenestrae by electron microscopy, and tracer behavior all matter. Its presence alone does not turn a capillary into an open passage for an intact bacterial cell.

PV1 immunohistochemistry should be evaluated alongside endothelial markers such as CD31 or CD34, the intestinal segment selected, the microvessel type, and staining controls. The number of PV1-positive pixels may change because of angiogenesis, an increase in vessel area, or a change in vessel composition—not only because flow through each vessel has changed. A causal conclusion requires manipulation of the candidate factor, measurement of barrier flux, and exclusion of alternative explanations. The source manuscript’s claim that “PV1 = 0 in normal tissue” and that any detection means permeability to particles larger than 70 kDa is too categorical for a heterogeneous intestinal microvascular network.

The source text proposed measuring soluble PLVAP in serum, with a threshold of 2 ng/mL and a prediction of mortality in cirrhosis. The cited paper concerns a bacterial protease, not GVB. We therefore do not present sPLVAP as a validated test of intestinal vascular permeability, report a threshold, or suggest tracking “fenestral closure” with it. Tissue PV1 may be informative in research; translation to clinical serum testing requires separate analytical and prognostic validation.

## 3. Regulation: Wnt/β-catenin, inflammation, and vascular plasticity

### Endothelial Wnt signaling in specific models

Canonical Wnt signaling begins when a ligand binds Frizzled and LRP5/6, attenuates β-catenin destruction by the Axin–APC–GSK3β complex, and allows β-catenin to participate in the TCF/LEF nuclear program. This pathway is important to the barrier phenotype during brain-vessel development; Spadoni and Mouries provided direct evidence for its role in the intestinal endothelium of animals. In a high-fat diet and another mouse model, endothelial β-catenin activation by Mouries prevented some GVB changes and features of steatohepatitis; samples from the colons of patients with NASH showed a marker of suspected vascular leakage. These evidence levels must remain distinct: genetic endothelial activation in a mouse is a causal experiment in a model, while staining human tissue is an association.

The source manuscript turns β-catenin into a simple switch: “Wnt on → PLVAP strictly off → sealed barrier.” β-catenin is also part of the VE-cadherin adherens-junction complex, and its nuclear and membrane pools have different functions. Wnt effects on PLVAP, CLDN5, angiogenesis, and vascular identity depend on the vascular bed and cell state. Without a separate promoter experiment, one should not claim that TCF/LEF directly binds the *PLVAP* promoter in intestinal endothelial cells and represses it under all conditions. That proposed link is not necessary to describe the demonstrated protective effect of endothelial β-catenin in a mouse model.

### VEGF, Ang/Tie2, cytokines, and metabolic stress

VEGF-A can increase vascular permeability through VEGFR2 by activating Src-dependent events and internalization of VE-cadherin. In endothelial-cell experiments, β-arrestin-dependent endocytosis of VE-cadherin has been shown as a mechanism of the VEGF response. But finding VEGF in inflamed intestine does not prove that every GVB endothelial defect follows the same Tyr658 or Tyr731 residue; phosphorylation, protein localization, and flux must be measured in the same model. Angiopoietin-1/Tie2 often stabilizes junctions in vascular biology, whereas Ang-2 acts according to inflammatory context and receptor availability. It is useful to describe these axes as a possible regulatory framework, but they do not yet support giving Ang-1 or a VEGF inhibitor to “seal the GVB” in people.

TNF-α, IL-1β, and bacterial ligands can alter the cytoskeleton, leukocyte adhesion, and endothelial function. Leukocyte passage through a vessel wall is a regulated process and is not the same as free passage of bacterial particles. During ischemia, HIF-1α is stabilized when oxygen-dependent hydroxylation and VHL recognition decline; the HIF program can change VEGF, metabolism, and tissue survival. During oxidative stress, BH4 availability may affect eNOS coupling and the NO/superoxide balance. These biochemical pathways are established beyond the intestinal vascular field, but the source text connects them to specific “GVB breaks” and treatments without direct confirmation. They are not depicted here as an obligatory sequence of events.

### Flow direction and cytoskeletal biochemistry

In endothelial cells, VEGFR2, Tie2, TLR receptors, and mechanosensitive complexes integrate growth, inflammatory, and flow signals. Phosphatases remove phosphate groups from receptors and cadherin complexes, while kinases add them; the outcome depends on enzyme localization and stimulus duration. Increased actomyosin tension through RhoA/ROCK or MLCK can pull cell edges apart and widen interendothelial gaps, whereas Rac1-dependent spreading stabilizes junctions in some contexts. These mechanisms, however, were derived from different vessels and cell cultures. For GVB, specifying the sequence “VEGF → Src → one tyrosine → bacteria in the portal vein” without measuring each intermediate would overextend the evidence.

The endothelial glycocalyx contacts blood, not intestinal contents. This distinction matters for claims about bacteria: a bacterium in the lumen must already cross the epithelium, interstitium, and basal side of the vessel wall—or enter the vessel by another route—to reach that layer. Loss of glycocalyx during systemic inflammation does not directly show that a dietary component opened a path **from the lumen** into the portal vein. Models should specify the direction of exposure: the apical side of an enterocyte, the abluminal side of the endothelium, and the luminal side of a blood vessel are different surfaces.

### *Salmonella*: a checked genetic result

The primary Spadoni study links *S. Typhimurium* passage through the GVB to the type III secretion system encoded by pathogenicity island **SPI-2** and to reduced β-catenin-dependent signaling in intestinal endothelium. It does not establish the source text’s proposed specific mechanism in which “SPI-1/AvrA acetylates the pathway, and ΔavrA cannot reach the liver.” AvrA may participate in other cellular processes of *Salmonella*, but a specific role for this effector in GVB requires direct comparison of isogenic strains and measurement of the endothelium. This example shows why a microbial genetic effect should not be replaced by an attractive but unsupported chain of molecular arrows.

## 4. When the vascular route is linked to the liver

### 4.1 Portal products and the immune response

After disruption of the mucosal or vascular boundary, bacterial components may enter portal blood. LPS binds LBP and is transferred to CD14/MD-2/TLR4; the intracellular MyD88 and TRIF branches initiate inflammatory and interferon responses. Kupffer cells take up some bacterial material, while hepatic stellate cells can respond to chronic injury and profibrotic signals. The experiment by Seki and colleagues linked TLR4 to increased TGF-β signaling and fibrosis in the relevant model. But a high peripheral serum LBP level does not identify the route by which a product left the intestine and does not measure GVB. Portal clearance and systemic inflammation also depend on liver status, microbial burden, the epithelial barrier, and immune function. For the separate role of intestinal alkaline phosphatase in LPS handling, see [Intestinal alkaline phosphatase: LPS chemistry, the microbiota, and clinical limits](/knowledge/intestinal-alkaline-phosphatase-iap-lps-detoxification).

Scale matters in vascular diagrams, too. A fenestra may pass a soluble complex of a given size; a viable bacterium is much larger than a molecular tracer. Live bacteremia therefore cannot be inferred from a single PV1-positive vessel or increased dextran flux. Studies should separately measure bacterial colony-forming units or reliable alternatives, rule out contamination in low-biomass samples, and account for hepatic bacterial clearance. “Portal endotoxemia” is meaningful only when the measurement and controls are clearly defined, not as an explanation for any elevated aminotransferase.

The molecular sequence LPS→LBP/CD14→TLR4/MD-2→MyD88/TRIF→NF-κB describes recognition; it is not a direct measure of fibrosis severity. A profibrotic response requires persistent injury, changes in stellate cells, a balance between extracellular-matrix formation and breakdown, and time. In some tissues bacterial DNA may be detected while viable culture is negative; these are different biological findings. Reagent controls, collection procedures, and comparison with background are essential when microbial biomass is low. These limits matter especially when making a strong claim about a “sterile” or “infected” portal compartment.

### 4.2 Steatohepatitis, alcohol-related disease, and cirrhosis

Mouries and colleagues studied mouse models of NAFLD/NASH: epithelial and GVB changes appeared early, microbiota transfer from high-fat-diet animals induced changes in a vascular marker in recipients, and endothelial β-catenin activation prevented some outcomes. Obeticholic acid acted within the same experimental design. In humans, the authors described an increased marker of vascular leakage in the colon of patients with NASH, but did not provide randomized evidence that the drug treats disease through GVB. Current clinical decisions in MASH follow validated guidance on disease stage and treatment, not a separate “PV1 protocol.”

Alcohol-related liver disease, cirrhosis, and spontaneous bacterial peritonitis illustrate the importance of the gut microbiota, epithelium, immune response, and portal circulation. GVB may be one boundary under study, but portal hypertension, bacterial overgrowth, motility, immune dysfunction, and hepatic clearance make it difficult to declare GVB the sole cause. Systematic reviews of bacterial translocation in cirrhosis do not validate the source manuscript’s serum sPLVAP or one LBP threshold for all patients. Distinguishing liver-disease type and stage is more useful than grouping MASH, alcohol-related injury, and cirrhosis into a single “GVB breakdown syndrome.”

### 4.3 Colorectal cancer and the hepatic pre-metastatic niche

Bertocchi and colleagues showed that *E. coli* in primary colorectal tumors can disrupt GVB in their experimental system, and bacterial spread to the liver was associated with an environment favorable to metastases. The primary paper emphasized the virulence regulator **VirF**. This is an important causal observation in a model and a tissue context in human cancer; it does not mean that elevated PV1 causes metastases in every patient with CRC or that a probiotic will prevent them. Metastasis depends on tumor cells, vascular invasion, immune and hepatic factors, treatment, and follow-up time. Clinical prediction would require independent validation of a GVB measurement, added value beyond cancer stage, and an intervention trial with metastasis as an outcome.

[![Four distinct GVB evidence blocks: a mouse Salmonella model, mice and human tissue in MASH, tumor-associated bacteria in CRC, and requirements for clinical validation.](https://api.nutrifit.health/api/v2/public/knowledge/files/images/nutrifit-scientific-20261009-gut-vascular-barrier-gvb-pv1-liver-metastasis-leaks-gvb-evidence-bridge-523bd3d70634.png)](https://api.nutrifit.health/api/v2/public/knowledge/files/images/nutrifit-scientific-20261009-gut-vascular-barrier-gvb-pv1-liver-metastasis-leaks-gvb-evidence-bridge-523bd3d70634.png)

The genotoxin colibactin produced by some *pks*⁺ *E. coli* can damage epithelial-cell DNA in experiments. This is a separate carcinogenesis pathway, not the same as VirF-dependent GVB disruption and not an obligatory cause of every metastatic niche. Combining both into one chain—“a *pks*⁺ bacterium opens PV1 and triggers metastasis”—mixes different strains, mechanisms, and stages of tumor disease.

An experimental pre-metastatic niche implies a sequence: a microbe leaves the tumor region, reaches the liver, causes local immune or stromal changes, and then tumor cells can establish themselves more easily. Each transition can be tested separately. Finding a bacterium in a primary tumor does not prove delivery to the liver; a bacterium in the liver does not show that it preceded a metastasis; and co-occurrence of PV1 and metastases does not establish causal direction. Animal studies can intervene on bacteria or vascular signaling and observe outcomes, whereas human series more often describe co-occurrence in established disease. Without this distinction, a compelling biological model becomes an unsupported clinical test.

### 4.4 Other diseases and limits of extrapolation

The original study described GVB changes in patients with celiac disease and elevated aminotransferases. This is a limited association in a specific group, not evidence that GVB causes celiac disease, primary sclerosing cholangitis, or any systemic autoimmune disease. A study of the brain choroid-plexus vascular barrier during intestinal inflammation concerns a different anatomical boundary; it cannot be recast as proof that intestinal PV1 opens the blood–brain barrier. Inflammatory bowel disease, infection, and sepsis can alter endothelium, but the severity and location of inflammation, treatment, and comorbidities shape interpretation. Increased PV1 in inflamed tissue alone does not identify the causal step in disease.

### 4.5 IBD, short bowel, and vascular remodeling

In Crohn’s disease and ulcerative colitis, inflammation affects more than epithelial junctions: endothelial activation, leukocyte recruitment, and angiogenesis also change. These processes make GVB a useful research subject, but they are not equivalent to one molecular “leak.” Anti-TNF and other medicines may reduce inflammation and, consequently, vascular markers. To claim that treatment works specifically through GVB, investigators must show temporal order, a separate endothelial measurement, and independence from a simple reduction in epithelial inflammation. The absence of a validated clinical “GVB score” does not prevent endothelial research, but it does preclude using such a score to decide when to change treatment.

After small-bowel resection, the remaining intestine adapts: villi, blood flow, nutrient transport, and microbial conditions change. Vascular remodeling may occur, but no universal PV1 value determines adaptation success or the need for parenteral nutrition. Outcomes in short-bowel syndrome depend primarily on the length and anatomy of the remnant, whether colon remains, nutrition, and complications. The simple chain “short bowel → open GVB → portal endotoxemia” does not account for this heterogeneity. Patient comparisons should match surgical anatomy and inflammation before vascular staining is interpreted.

### 4.6 The hepatic and brain barriers do not directly extend the GVB

The hepatic sinusoid has its own fenestrated endothelium and powerful immune system. A microbial product entering portal blood does not necessarily circulate freely in peripheral blood; the liver removes or transforms some of it. Cirrhosis changes clearance, and portal pressure and collaterals can redirect flow. Comparing portal and systemic concentrations without accounting for liver status can therefore mislead about intestinal vascular permeability. Changes in aminotransferases also cannot locate the first damaged boundary.

The brain choroid plexus, blood–brain barrier, and intestinal microvessels have different endothelial programs. Carloni and colleagues’ study of choroid-plexus barrier closure during intestinal inflammation raises a question about inter-organ regulation; it does not confirm the chain “intestinal PV1 → bacteria in the brain → neurological disease.” Such a conclusion would require tracking the same material along the route, measuring both barriers, and assessing a neurological outcome. Associations between intestinal inflammation and central nervous system symptoms are not enough by themselves.

## 5. How to study GVB without substituting tests of neighboring systems

A research protocol should state the tracer’s direction of movement, size, route of administration, vascular segment, and time after injury. If dextran is put into the intestinal lumen and later found in blood, the result combines epithelial, tissue, endothelial, and clearance pathways. An intravascular injection followed by observation of extravasation tests a different direction. PV1 staining alongside CD31/CD34 describes vascular phenotype; electron microscopy shows fenestrae and diaphragms; bacterial cultures or sequencing require rigorous contamination controls. A microfluidic gut-vascular-on-a-chip model controls flow and coculture, but does not reproduce all villus anatomy, immunity, or the portal liver.

The lactulose/mannitol test primarily assesses passage across intestinal epithelium and renal excretion under a defined protocol. Fecal calprotectin indicates neutrophilic inflammation, not the state of the vessel wall. I-FABP may reflect enterocyte injury. LBP and sCD14 relate to responses to microbial products but depend on inflammation and liver function. Fibroelastography and CAP help assess certain properties of the liver; they are not a window onto GVB. The studies reviewed do not establish a clinically validated serum PV1 threshold, marker combination, or “GVB recovery in 8–12 weeks.”

| Measurement | What it most directly reflects | Main limitation |
| :--- | :--- | :--- |
| Tissue PV1 + endothelial marker | Phenotype of the sampled vascular segment | Not quantitative bacterial flux |
| Dextrans and other tracers | Permeability for a particular size and route | Not universal impermeability to all particles |
| Portal/hepatic microbiology | Bacterial spread in a specific model | Contamination risk and hepatic clearance |
| Lactulose/mannitol | Epithelial test under a defined protocol | Not specific to GVB |
| LBP, sCD14, calprotectin | Response to microbial products or inflammation | Does not localize an endothelial defect |
| Aminotransferases, CAP, elastography | Liver injury and properties | Does not show the route by which antigen entered |

### Limits of a vascular biomarker

Showing a difference in average staining between groups is not enough to make PV1 clinically useful. Biopsy and tissue processing must be standardized; within- and between-laboratory reproducibility, dependence on segment and vessel count, sensitivity and specificity in an independent cohort, and usefulness for a decision that changes patient outcomes must be assessed. Serum analysis would additionally require validation of the molecular form and protein source, half-life, effects of liver and kidney function, and agreement with tissue GVB. An arbitrary fall in a marker after a supplement could reflect another inflammatory process.

Even a validated prognostic marker is not necessarily a therapeutic target. If PV1 rises in response to vascular injury, suppressing it may not repair the vessel wall; in some fenestrated vessels, PV1 diaphragms help limit protein escape. Genetic and pharmacological experiments can distinguish marker from mechanism by measuring not only staining but also vessel function, immune protection, and adverse effects. A clinical test should measure what a clinician needs to know, not merely a convenient laboratory protein.

## 6. What treatment can and cannot promise

Protecting the epithelium, treating infection, and managing underlying liver disease are clinically meaningful goals, but specific “GVB repair” remains primarily a research objective. The Mouries study found that endothelial β-catenin intervention and obeticholic acid acted in a mouse model; translation to people requires evaluation of efficacy and safety against clinical outcomes. Raising butyrate through microbial fermentation may support epithelial metabolism and immune regulation, but epithelial data do not establish direct induction of Tie2 in intestinal capillaries or closure of PV1-positive fenestrae. Indole metabolites and AhR, polyphenols, omega-3 fatty acids, and inflammation-resolution mediators have plausible biological pathways, but the cited sources do not support GVB dosing regimens, “92% sealing” by a combination of butyrate and curcumin, or metastasis prevention.

Supplement formulations do not remove the need to measure tissue exposure. Improved solubility in a phytosome does not prove delivery of unchanged polyphenol to a pericryptal vessel at the needed concentration; a butyrate capsule does not guarantee a defined level near the endothelium. The source manuscript proposed probiotic consortia, pasteurized *Akkermansia*, UDCA, vitamin E, coenzyme Q10, alpha-lipoic acid, and high-dose EPA/DHA as one “GVB protocol.” *Akkermansia* studies concerned metabolic outcomes in mice and a small human study, not proven GVB repair. For strain-specific effects and safety considerations, see this [overview of probiotic strains, clinical effects, and safety](/knowledge/probiotics-gut-microbiome-safety); it does not demonstrate GVB repair. Medicines and nutraceuticals have their own indications, contraindications, and risks; prescribing them for an unvalidated “PV1 syndrome” is unsupported.

A practical research path is more specific than a universal supplement plan: select a disease and population; measure epithelium and vascular endothelium separately; identify the microbial material and its route; test an intervention first against a model vascular outcome and then a human clinical outcome. In oncology, that outcome would be confirmed metastasis and survival, not PV1 staining alone. In MASH, it would be validated liver outcomes together with vascular measurements, not a change in an arbitrarily selected serum protein.

### 6.1 Priorities for human research

The most informative design separates tissue measurements from clinical outcomes in advance. For example, elective colorectal surgery could provide standardized sampling of unaffected mucosa and tumor tissue, PV1 staining alongside endothelial markers, vascular-network quality, microbiological controls, and clinical stage. Prospective follow-up could then test whether the vascular measure adds prognostic information beyond established cancer factors. Intervention based on that marker would still require a separate trial: a prognostic marker may not be a modifiable cause.

For MASH, independent confirmation of the vascular signal across intestinal segments, control for diet, obesity, diabetes, and fibrosis stage, and testing whether GVB changes precede clinical progression would be important. Analysis of patients with severe disease alone cannot distinguish cause from consequence: liver dysfunction itself changes intestinal hemodynamics, immunity, and microbiota. Genetic or drug manipulation of endothelium in mice addresses part of the causal question, but human safety and tissue specificity remain separate questions.

## 7. The boundary of the evidence

GVB is a useful concept for the second tissue boundary between mucosa and portal blood. Primary experiments show that bacterial infection, microbiota-dependent conditions in steatohepatitis models, and tumor-associated bacteria can alter intestinal endothelium and the spread of microbial material. PV1 helps describe vascular phenotype, but does not replace direct measurement of flux or clinical outcomes. Human tissue associations in celiac disease, NASH, and cancer make this an important research direction; they do not establish a ready-to-use blood test, a nutraceutical dosing protocol, or a proven way to prevent metastases by “closing GVB.”

## FAQ

### What is the GVB, and how does it differ from the epithelial barrier?

The GVB is the vascular endothelial boundary between the intestinal lamina propria and blood. The epithelium separates the lumen from tissue, so impairment of one layer does not prove impairment of the other.

### Does positive PV1 staining mean bacteria entered the portal vein?

No. PV1 describes endothelial phenotype and diaphragmed structures. Bacterial movement requires a separate measurement, and staining depends on vessel and tissue type.

### Can a blood PLVAP test diagnose a “leaky GVB”?

The sources reviewed provide no clinically validated serum PLVAP threshold or protocol for tracking GVB closure.

### Is GVB proven to cause MASH?

GVB impairment contributes to steatohepatitis phenotypes in mouse models. Human tissue association does not show it is the sole cause of MASH or that GVB-directed treatment is already effective.

### Can PV1 predict colorectal-cancer metastases?

An experimental study links tumor-associated bacteria and GVB to a hepatic niche. PV1’s clinical prognostic accuracy and the benefit of treatment based on this marker have not been established.

### Should I take butyrate or polyphenols to repair GVB?

No regimen, dose, or clinical outcome for these supplements has been established for specific GVB repair in people. Treatment depends on the particular disease.

## Sources

- [Spadoni I, Zagato E, Bertocchi A, et al. A gut-vascular barrier controls the systemic dissemination of bacteria. Science. 2015;350(6262):830-834.](https://pubmed.ncbi.nlm.nih.gov/26564856/)
- [Bertocchi A, Carloni S, Ravenda PS, et al. Gut vascular barrier impairment leads to intestinal bacteria dissemination and colorectal cancer metastasis to liver. Cancer Cell. 2021;39(5):708-724.e11.](https://pubmed.ncbi.nlm.nih.gov/33798472/)
- [Mouries J, Brescia P, Silvestri A, et al. Microbiota-driven gut vascular barrier disruption is a prerequisite for non-alcoholic steatohepatitis development. J Hepatol. 2019;71(6):1216-1228.](https://pubmed.ncbi.nlm.nih.gov/31419514/)
- [Stan RV, Kubitza M, Palade GE. PV1 is a component of the diaphragms of fenestrae and stomata of caveolae in capillary endothelium. *Proc Natl Acad Sci U S A*. 1999;96(23):13203-13207.](https://pubmed.ncbi.nlm.nih.gov/10557299/)
- [Stan RV, Tse D, Deharvengt SJ, et al. The diaphragms of fenestrated endothelia: Gatekeepers of vascular permeability and transendothelial transport. *Am J Pathol*. 2012;181(5):1506-1518.](https://pubmed.ncbi.nlm.nih.gov/22960068/)
- [Liebner S, Corada M, Bangsow T, et al. Wnt/beta-catenin signaling controls development of the blood-brain barrier. *J Cell Biol*. 2008;183(3):409-417.](https://pubmed.ncbi.nlm.nih.gov/18955513/)
- [Daneman R, Agalliu D, Zhou L, Kuhnert F, Kuo CJ, Barres BA. Wnt/beta-catenin signaling is required for CNS, but not non-CNS, angiogenesis. *Proc Natl Acad Sci U S A*. 2009;106(2):641-646.](https://pubmed.ncbi.nlm.nih.gov/19129494/)
- [Vestweber D. VE-cadherin: the major endothelial adhesion molecule controlling cellular junctions and blood vessel formation. *Arterioscler Thromb Vasc Biol*. 2008;28(2):223-232.](https://pubmed.ncbi.nlm.nih.gov/18174452/)
- [Dejana E, Orsenigo F, Lampugnani MG. The role of adherens junctions and VE-cadherin in the control of vascular permeability. *J Cell Sci*. 2008;121(Pt 13):2115-2122.](https://pubmed.ncbi.nlm.nih.gov/18565824/)
- [Gavard J, Gutkind JS. VEGF controls endothelial-cell permeability by promoting the beta-arrestin-dependent endocytosis of VE-cadherin. *Nat Cell Biol*. 2006;8(11):1223-1234.](https://pubmed.ncbi.nlm.nih.gov/17060906/)
- [Savidge TC, Newman P, Pothoulakis C, et al. Enteric glia regulate intestinal barrier function and protect against colitis through S-nitrosoglutathione. *Gastroenterology*. 2007;132(4):1344-1358.](https://pubmed.ncbi.nlm.nih.gov/17408648/)
- [Armulik A, Genové G, Mäe M, et al. Pericytes regulate the blood-brain barrier. *Nature*. 2010;468(7323):557-561.](https://pubmed.ncbi.nlm.nih.gov/20944627/)
- [Albillos A, de Gottardi A, Rescigno M. The gut-liver axis in liver disease: Pathophysiological basis for therapy. *J Hepatol*. 2020;72(3):558-577.](https://pubmed.ncbi.nlm.nih.gov/31770592/)
- [Seki E, De Minicis S, Osterreicher CH, et al. TLR4 enhances TGF-beta signaling and promotes hepatic fibrosis. *Nat Med*. 2007;13(11):1324-1332.](https://pubmed.ncbi.nlm.nih.gov/17952090/)
- [Arthur JC, Perez-Chanona E, Mühlbauer M, et al. Intestinal inflammation targets cancer-inducing activity of the microbiota. *Science*. 2012;338(6103):120-123.](https://pubmed.ncbi.nlm.nih.gov/22903521/)
- [Nougayrède JP, Homburg S, Taieb F, et al. Escherichia coli induces DNA double-strand breaks in eukaryotic cells. *Science*. 2006;313(5788):848-851.](https://pubmed.ncbi.nlm.nih.gov/16902142/)
- [Dalmasso G, Cougnoux A, Delmas J, Darfeuille-Michaud A, Bonnet R. The bacterial genotoxin colibactin promotes colon tumor growth by inducing a senescence-associated secretory phenotype. *Sci Signal*. 2014;7(346):ra100.](https://pubmed.ncbi.nlm.nih.gov/25293671/)
- [Carloni S, Bertocchi A, Mancinelli S, et al. Identification of a choroid plexus vascular barrier closing during intestinal inflammation. *Science*. 2021;374(6566):439-448.](https://pubmed.ncbi.nlm.nih.gov/34672740/)
- [Rutkowski MR, Stephen TL, Svoronos N, et al. Microbially driven TLR5-dependent signaling governs distal malignant progression through tumor-promoting inflammation. *Cancer Cell*. 2015;27(1):27-40.](https://pubmed.ncbi.nlm.nih.gov/25500179/)
- [Kelly CJ, Zheng L, Campbell EL, et al. Crosstalk between Microbiota-Derived Short-Chain Fatty Acids and Intestinal Epithelial HIF Augments Tissue Barrier Function. *Cell Host Microbe*. 2015;17(5):662-671.](https://pubmed.ncbi.nlm.nih.gov/25865369/)
- [Zelante T, Iannitti RG, Cunha C, et al. Tryptophan catabolites from microbiota engage aryl hydrocarbon receptor and balance mucosal reactivity via interleukin-22. *Immunity*. 2013;39(2):372-385.](https://pubmed.ncbi.nlm.nih.gov/23973224/)
- [Plovier H, Everard A, Druart C, et al. A purified membrane protein from Akkermansia muciniphila or the pasteurized bacterium improves metabolism in obese and diabetic mice. *Nat Med*. 2017;23(1):107-113.](https://pubmed.ncbi.nlm.nih.gov/27892954/)
- [Depommier C, Everard A, Druart C, et al. Supplementation with Akkermansia muciniphila in overweight and obese human volunteers: a proof-of-concept exploratory study. *Nat Med*. 2019;25(7):1096-1103.](https://pubmed.ncbi.nlm.nih.gov/31263284/)
- [Cani PD, Bibiloni R, Knauf C, et al. Changes in gut microbiota control metabolic endotoxemia-induced inflammation in high-fat diet-induced obesity and diabetes in mice. *Diabetes*. 2008;57(6):1470-1481.](https://pubmed.ncbi.nlm.nih.gov/18305141/)
- [Brandl K, Hartmann P, Jih LJ, et al. Dysbiosis-induced intestinal inflammation activates tumor necrosis factor receptor I and mediates alcoholic liver disease in mice. *Hepatology*. 2017;65(3):855-868.](https://pubmed.ncbi.nlm.nih.gov/27862217/)
- [Wiest R, Lawson M, Geuking M. Pathological bacterial translocation in liver cirrhosis. *J Hepatol*. 2014;60(1):197-209.](https://pubmed.ncbi.nlm.nih.gov/23993913/)