Claudin5 protects the peripheral endothelial barrier in an organ and vessel-type-specific manner

  1. Mark Richards
  2. Emmanuel Nwadozi
  3. Sagnik Pal
  4. Pernilla Martinsson
  5. Mika Kaakinen
  6. Marleen Gloger
  7. Elin Sjöberg
  8. Katarzyna Koltowska
  9. Christer Betsholtz
  10. Lauri Eklund
  11. Sofia Nordling
  12. Lena Claesson-Welsh

  • Curated by eLife

    eLife logo

    Evaluation Summary:

    This manuscript provides new knowledge of one of the major molecules, claudin5, expected to maintain the integrity of the cell-cell contacts of the blood vessel wall. The results highlight organ and vessel-type specific significance of this mechanism in the regulation of vascular permeability, partially challenging the current view. A combination of in vivo microscopy and genetic mouse models is used to support the key claims in the paper. This manuscript will be of interest to scientists across vascular biology, and especially in the field of vascular permeability regulation.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. Reviewer #1 and Reviewer #2 agreed to share their name with the authors.)

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

Dysfunctional and leaky blood vessels resulting from disruption of the endothelial cell (EC) barrier accompanies numerous diseases. The EC barrier is established through endothelial cell tight and adherens junctions. However, the expression pattern and precise contribution of different junctional proteins to the EC barrier is poorly understood. Here, we focus on organs with continuous endothelium to identify structural and functional in vivo characteristics of the EC barrier. Assembly of multiple single-cell RNAseq datasets into a single integrated database revealed the variability and commonalities of EC barrier patterning. Across tissues, Claudin5 exhibited diminishing expression along the arteriovenous axis, correlating with EC barrier integrity. Functional analysis identified tissue-specific differences in leakage properties and response to the leakage agonist histamine. Loss of Claudin5 enhanced histamine-induced leakage in an organotypic and vessel type-specific manner in an inducible, EC-specific, knock-out mouse. Mechanistically, Claudin5 loss left junction ultrastructure unaffected but altered its composition, with concomitant loss of zonula occludens-1 and upregulation of VE-Cadherin expression. These findings uncover the organ-specific organisation of the EC barrier and distinct importance of Claudin5 in different vascular beds, providing insights to modify EC barrier stability in a targeted, organ-specific manner.

Article activity feed

  1. Version published to 10.7554/elife.78517 on eLife
  2. eLife

    Evaluation Summary:

    This manuscript provides new knowledge of one of the major molecules, claudin5, expected to maintain the integrity of the cell-cell contacts of the blood vessel wall. The results highlight organ and vessel-type specific significance of this mechanism in the regulation of vascular permeability, partially challenging the current view. A combination of in vivo microscopy and genetic mouse models is used to support the key claims in the paper. This manuscript will be of interest to scientists across vascular biology, and especially in the field of vascular permeability regulation.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. Reviewer #1 and Reviewer #2 agreed to share their name with the authors.)

  3. eLife

    Reviewer #1 (Public Review):

    Abnormal vascular leakage is one of the hallmarks of vascular injury and inflammation, which are frequently accompanied by major diseases such as diabetes, sepsis, and wet-AMD. It is induced by the impairment of the endothelial cell (EC) barrier. The integrity of the EC barrier is accomplished primarily by proper arrangements of adherens junctions (AJs) and tight junctions (TJs), where transmembrane proteins form intercellular interactions to bridge adjacent EC membranes. Among these, claudin5 is known to be a key molecule in maintaining blood-brain-barrier (BBB), but its function in the peripheral tissues has been poorly understood. In this study, throughout the assembly of multiple single‐cell RNAseq datasets into a single integrated database, the authors found that claudin5 expression diminishes along the arteriovenous axis, which correlates with EC barrier integrity. They elegantly showed that the claudin5 deficiency enhances histamine‐induced leakage in organ- and vessel type‐specific, and size‐selective manners. They reasoned that these could be the result of alternative compositions of AJs and TJs, and this study will aid our ability to modify EC barrier stability in a targeted, organ‐specific manner. Overall, the data are analyzed thoroughly and the conclusions drawn are novel and appealing. Understanding molecular and cellular mechanisms for EC barrier integrity at homeostasis and its impairment at pathologic conditions are fundamental and central topics for understanding vascular biology and related cardiovascular diseases, making this study timely and important. It would be constructive if the authors provide more exquisite underlying molecular and cellular mechanisms to support their claims.

  4. eLife

    Reviewer #2 (Public Review):

    Richards et al. investigate the function of the tight junction protein Claudin-5 (Cldn5) in the regulation of permeability of the mature vasculature, with a focus on tissues with continuous capillary cell-cell junctions. Richards et al. employ an inducible endothelial cell-specific deletion of Cldn5 in adult mice, which enabled challenging the mice using histamine. Previously, such studies were not possible due to neonatal lethality of the mice where Cldn5 was deleted constitutively leading to loss of the blood-brain barrier (BBB) integrity. Richards et al. show that Cldn5 has an organ-specific and vessel-type specific function in the regulation of size-selective vascular permeability in response to histamine (>2000 kDa), whereas basal permeability (to 4-70kDa substances) was not affected. In line with this, loss of Cldn5 did not cause severe disruption of the ultrastructure of the endothelial cell-cell junctions but made the junctions more permeable in response to challenge. Loss of Cldn5 in vivo increased the expression of VE-Cadherin and decreased zonula occludens‐1, proteins associated with adherens and tight junctions, respectively, which may have affected the observed vascular phenotype.

    The authors have previously shown that vascular segments with little or no expression of Cldn5 are most permeable in the mouse ear dermis, where Cldn5 is most highly expressed in arterioles but decreased towards capillary and venous beds. Extending these studies to other tissues, including the back skin, trachea, skeletal muscle and the heart, Richards et al. found decreased Cldn5 expression along the arteriovenous axis, but not a similarly correlated permeability pattern, suggesting other mechanisms. In the ear dermis, Cldn5 deletion did not increase the permeability of arterioles but increased the number and extent of leakage sites in small and mid-sized vessels (up to 15 um) with low Clnd5 expression.

    The conclusions by Richards et al. are mostly well supported by data. The results to some extent challenge the concept of the significance of Cldn5 in vascular integrity and indicate that the function of Cldn5 is highly organ and vessel-type specific. Compensatory mechanisms by other junction proteins can not be excluded. Compared to the known function of Cldn5, the permeability of the CNS vasculature to small substances was not investigated, and thus the potential differential sensitivity of the mature vasculature to the loss of Cldn5 is not considered.

  5. eLife

    Reviewer #3 (Public Review):

    This manuscript is of broad interest to readers who study vascular integrity that controls exchanging molecules across the endothelial cell and parenchymal cells. It uncovers role for inter-endothelial cell-cell adhesion molecules expressed in manners-dependent on blood vessels in vascular barrier function. The analyses using gene and protein expression and using imaging techniques support the claims of this manuscript.

    The authors first examined the expression of vessel-specific markers using scRNA-seq datasets and grouped them into 5 subsets: arterial, arterial/capillary, capillary, capillary/venous, and venous. Among the tight junction (TJ) molecules, they focused on claudin5, because its expression diminished along the arteriovenous axis. Vascular leakage outside of the central nervous system (CNS) including ear skin, back skin, skeletal muscle, trachea, and heart was tested in the mice conditionally depleted of cldn5 because they previously examined the correlation of Cldn5 expression with vessel diameter using Cldn(BAC)-GFP mice. Furthermore, basal vascular leakage and inflammatory cytokine (histamine)-induced leakage was quantitatively analyzed to investigate the relevance of Claudin5 expression to vascular integrity. Depletion of Claudin5 affected the leakage and resulted in variation of leakage that was dependent on vessel subtypes. Finally, they found no apparent junctional structure but noticed compensatory compositional changes of adhesion molecules in the Cldn5-deficient mice. Cladin5 expression inversely correlated with that of adherence junction (AJ) molecule, vascular endothelial cadherin (VE-Cadherin) and TJ molecule, Occludin, while its expression positively correlated with that of ZO-1.

    The present data support the authors' claim that Claudin5-dependent barrier function varied in the organ vessels and demonstrate one example of organotypic vessel structure and function in vivo. To precisely confirm their claim, there are several points that should be explained. Several experiments that strengthen their conclusions are required for clarification of their claim.

    Strength
    1. Gene expression of Inter-endothelial junction molecules including AJ molecules (Cadherin family members), TJ molecules (Claudin family genes, JAM, Occludin, Nectin, etc), and intracellular molecules (Amot, Cingulin, Catenin) was thoroughly and quantitatively examined in endothelial cell subsets.
    2. Spaciotemporal leakage of various sizes of molecules from endothelial subsets was clearly imaged to understand the relevance of Claudin5 expression to barrier function.
    3. To analyze the role of Claudin5 in barrier function, the authors used both Claudin (BAC)-GFP transgenic mice and Cldn5 iEC knockout mice.
    These methods enabled them to demonstrate the importance of Claudin5 for organotypic barrier function.

    Weakness
    1. Although barrier function is carefully investigated from the view of adhesion molecules, the expression pattern of cytokine receptors might be also involved in cytokine-induced permeability and leakage. If cytokine receptors are not evenly expressed in vessel subsets, the permeability might vary in arteriole, capillary, and venule.
    2. Gene expression is thoroughly investigated in the ear skin, back skin, trachea, skeletal muscle and heart. However, protein expression was not carefully examined. Claudin5 was still expressed even after the cldn5 gene was knocked out in endothelial cells. Therefore, immunohistochemistry of blood vessels should be carefully performed to confirm where the compensatory increase of gene expression is found.
    3. Given the dominant expression of Claudin5 in arterioles over venules, leakage in the cldn5 KO mouse should be mostly found in arterioles. The authors did not show clear data about the leakage sites. Figure 4B vi and Figure 3C ii suggested the leakage in the capillaries and venules.
    4. The interpretation of the results might need a more careful explanation.

    Overall, the results support their claim. However, additional experiments would fortify their claim and make their claims more logical.

  6. Version published to 10.1101/2022.03.14.484246v1 on bioRxiv