Limits...
Organ-wide 3D-imaging and topological analysis of the continuous microvascular network in a murine lymph node.

Kelch ID, Bogle G, Sands GB, Phillips AR, LeGrice IJ, Dunbar PR - Sci Rep (2015)

Bottom Line: By focussing on critical immune microenvironments we quantified differences in their vascular topology.We further developed a morphology-based approach to identify High Endothelial Venules, key sites for lymphocyte extravasation.These data represent a comprehensive and continuous blood vessel network of an entire organ and provide benchmark measurements that will inform modelling of blood vessel networks as well as enable comparison of vascular topology in different organs.

View Article: PubMed Central - PubMed

Affiliation: Maurice Wilkins Centre, University of Auckland, Private Bag 92-019, Auckland 1142, New Zealand.

ABSTRACT
Understanding of the microvasculature has previously been limited by the lack of methods capable of capturing and modelling complete vascular networks. We used novel imaging and computational techniques to establish the topology of the entire blood vessel network of a murine lymph node, combining 63,706 confocal images at 2 μm pixel resolution to cover a volume of 3.88 mm(3). Detailed measurements including the distribution of vessel diameters, branch counts, and identification of voids were subsequently re-visualised in 3D revealing regional specialisation within the network. By focussing on critical immune microenvironments we quantified differences in their vascular topology. We further developed a morphology-based approach to identify High Endothelial Venules, key sites for lymphocyte extravasation. These data represent a comprehensive and continuous blood vessel network of an entire organ and provide benchmark measurements that will inform modelling of blood vessel networks as well as enable comparison of vascular topology in different organs.

No MeSH data available.


Related in: MedlinePlus

Comparative analysis of LN subregions.Based on regional differences in vessel density, functionally distinct LN subregions were isolated and evaluated. LN subnetworks colour-coded by location (red = predicted T-cell region, blue = B-cell follicles, yellow = hilum region, green = dense vasculature) are shown together with the whole vascular topology map (grey) in Amira (a). The average distribution of the distances to the closest blood vessel in each region reveals the characteristic vascular coverage in different parts of the LN (b). The mean blood vessel distance in LN subnetworks shows strong similarities between the predicted T- and B-cell regions, while being generally enlarged near the hilum, and notably shorter in dense regions (c). Vessel densities were evaluated using a 2D histology tool and 3D counts by the topology program, which both revealed similar proportions between LN subnetworks (d). With the exception of the hilum, which contains a range of very large and small vessels, the average diameter within LN subnetworks was nearly constant (e). A close-up of a B-cell follicle shows the avascular appearance in a 2D section, the selected vascular subnetwork (blue) enclosing a void with a distance greater than 60 μm from the nearest blood vessel (red), and the local pHEVs (magenta). Measurements of the mean and standard deviation (b–e) are representative of 13 individual 3D blocks (with a dimension of approximately 200 × 200 × 200 μm) from the predicted T- and the B-cell regions, and 5 blocks from the hilum and dense regions, respectively. Data were analysed using 1-way ANOVA, Tukey’s comparison. ***P < 0.001. ns = not significant, P > 0.05. Fo = follicle. Scale bar: 100 μm. The locations of analysied LN subregions together with avascular voids are visualised in Supplementary Movie 5.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4645097&req=5

f7: Comparative analysis of LN subregions.Based on regional differences in vessel density, functionally distinct LN subregions were isolated and evaluated. LN subnetworks colour-coded by location (red = predicted T-cell region, blue = B-cell follicles, yellow = hilum region, green = dense vasculature) are shown together with the whole vascular topology map (grey) in Amira (a). The average distribution of the distances to the closest blood vessel in each region reveals the characteristic vascular coverage in different parts of the LN (b). The mean blood vessel distance in LN subnetworks shows strong similarities between the predicted T- and B-cell regions, while being generally enlarged near the hilum, and notably shorter in dense regions (c). Vessel densities were evaluated using a 2D histology tool and 3D counts by the topology program, which both revealed similar proportions between LN subnetworks (d). With the exception of the hilum, which contains a range of very large and small vessels, the average diameter within LN subnetworks was nearly constant (e). A close-up of a B-cell follicle shows the avascular appearance in a 2D section, the selected vascular subnetwork (blue) enclosing a void with a distance greater than 60 μm from the nearest blood vessel (red), and the local pHEVs (magenta). Measurements of the mean and standard deviation (b–e) are representative of 13 individual 3D blocks (with a dimension of approximately 200 × 200 × 200 μm) from the predicted T- and the B-cell regions, and 5 blocks from the hilum and dense regions, respectively. Data were analysed using 1-way ANOVA, Tukey’s comparison. ***P < 0.001. ns = not significant, P > 0.05. Fo = follicle. Scale bar: 100 μm. The locations of analysied LN subregions together with avascular voids are visualised in Supplementary Movie 5.

Mentions: We noticed clear regional differences in the vascular composition and density. Compared to locations at the LN border, the central region was characterised by more vessels with diameters of 20–30 μm, several voids with a distance greater than 60 μm to the nearest blood vessel, and a higher branch separation to the main vein. This region was found to correspond to the paracortical T-cell region of mesenteric LNs when investigated by 2D immunofluorescence microscopy (Supplementary Figure 2). As described above, regions with an almost avascular appearance located in the outer rim of the LN had typical morphological features of B-cell follicles. Based on these observations, we analysed LN subregions with visually discernible blood vascular morphology with respect to their diameter distribution, vessel distance, and vessel density. We selected 3D blocks from a central sphere (the predicted paracortical T-cell zone), the B-cell follicles (at the rim), and the hilum, and computed the distance to the nearest blood vessel (Fig. 7a,b). The predicted T- and B-cell regions show very similar distance distributions that are lower than those in the hilum, and slightly higher than the average distance within the whole network. This trend is also evident when the mean distances within subregions are compared (Fig. 7c). The equivalence of vessel distances in T and B cell regions initially seems counter-intuitive given the low density of vessels within the follicle centre. However, the analysis shown in Fig. 7 not only incorporates the follicular core but also the vasculature immediately surrounding follicles, which compensates for the high vessel distances that can be found within the centre of the follicles (Supplementary Figure 4). We also identified regions of dense vasculature outside these three defined zones, characterised by shorter distances to supplying blood vessels (Fig. 7a–c). To compare our 3D measurements with the conventional method for evaluating tissue sections, we created a 2D histology tool to evaluate vascular density in different compartments in 2D. This tool confirmed large density differences between subcompartments, and measurements of vascular density in 2D had the same hierarchy as in 3D (Fig. 7d). A combination of several different measurements that are feasible within the 3D model provides a new vision of the 3D topology of vessels within specific regions of the LN (Fig. 7f). Although B-cell follicles can appear nearly avascular in 2D images (Fig. 7f, upper panel), 3D visualisation shows the surrounding vascular loops supplying the follicle, as well as a central area within the follicle that is >60 μm distal from any of these vessels, which could be found in the majority of follicles in this study. Nearby HEVs that supply lymphocytes to the follicle can also be highlighted within the same image projections to provide a complete 3D view of the vascular environs (Fig. 7f) that is not evident in 2D views.


Organ-wide 3D-imaging and topological analysis of the continuous microvascular network in a murine lymph node.

Kelch ID, Bogle G, Sands GB, Phillips AR, LeGrice IJ, Dunbar PR - Sci Rep (2015)

Comparative analysis of LN subregions.Based on regional differences in vessel density, functionally distinct LN subregions were isolated and evaluated. LN subnetworks colour-coded by location (red = predicted T-cell region, blue = B-cell follicles, yellow = hilum region, green = dense vasculature) are shown together with the whole vascular topology map (grey) in Amira (a). The average distribution of the distances to the closest blood vessel in each region reveals the characteristic vascular coverage in different parts of the LN (b). The mean blood vessel distance in LN subnetworks shows strong similarities between the predicted T- and B-cell regions, while being generally enlarged near the hilum, and notably shorter in dense regions (c). Vessel densities were evaluated using a 2D histology tool and 3D counts by the topology program, which both revealed similar proportions between LN subnetworks (d). With the exception of the hilum, which contains a range of very large and small vessels, the average diameter within LN subnetworks was nearly constant (e). A close-up of a B-cell follicle shows the avascular appearance in a 2D section, the selected vascular subnetwork (blue) enclosing a void with a distance greater than 60 μm from the nearest blood vessel (red), and the local pHEVs (magenta). Measurements of the mean and standard deviation (b–e) are representative of 13 individual 3D blocks (with a dimension of approximately 200 × 200 × 200 μm) from the predicted T- and the B-cell regions, and 5 blocks from the hilum and dense regions, respectively. Data were analysed using 1-way ANOVA, Tukey’s comparison. ***P < 0.001. ns = not significant, P > 0.05. Fo = follicle. Scale bar: 100 μm. The locations of analysied LN subregions together with avascular voids are visualised in Supplementary Movie 5.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4645097&req=5

f7: Comparative analysis of LN subregions.Based on regional differences in vessel density, functionally distinct LN subregions were isolated and evaluated. LN subnetworks colour-coded by location (red = predicted T-cell region, blue = B-cell follicles, yellow = hilum region, green = dense vasculature) are shown together with the whole vascular topology map (grey) in Amira (a). The average distribution of the distances to the closest blood vessel in each region reveals the characteristic vascular coverage in different parts of the LN (b). The mean blood vessel distance in LN subnetworks shows strong similarities between the predicted T- and B-cell regions, while being generally enlarged near the hilum, and notably shorter in dense regions (c). Vessel densities were evaluated using a 2D histology tool and 3D counts by the topology program, which both revealed similar proportions between LN subnetworks (d). With the exception of the hilum, which contains a range of very large and small vessels, the average diameter within LN subnetworks was nearly constant (e). A close-up of a B-cell follicle shows the avascular appearance in a 2D section, the selected vascular subnetwork (blue) enclosing a void with a distance greater than 60 μm from the nearest blood vessel (red), and the local pHEVs (magenta). Measurements of the mean and standard deviation (b–e) are representative of 13 individual 3D blocks (with a dimension of approximately 200 × 200 × 200 μm) from the predicted T- and the B-cell regions, and 5 blocks from the hilum and dense regions, respectively. Data were analysed using 1-way ANOVA, Tukey’s comparison. ***P < 0.001. ns = not significant, P > 0.05. Fo = follicle. Scale bar: 100 μm. The locations of analysied LN subregions together with avascular voids are visualised in Supplementary Movie 5.
Mentions: We noticed clear regional differences in the vascular composition and density. Compared to locations at the LN border, the central region was characterised by more vessels with diameters of 20–30 μm, several voids with a distance greater than 60 μm to the nearest blood vessel, and a higher branch separation to the main vein. This region was found to correspond to the paracortical T-cell region of mesenteric LNs when investigated by 2D immunofluorescence microscopy (Supplementary Figure 2). As described above, regions with an almost avascular appearance located in the outer rim of the LN had typical morphological features of B-cell follicles. Based on these observations, we analysed LN subregions with visually discernible blood vascular morphology with respect to their diameter distribution, vessel distance, and vessel density. We selected 3D blocks from a central sphere (the predicted paracortical T-cell zone), the B-cell follicles (at the rim), and the hilum, and computed the distance to the nearest blood vessel (Fig. 7a,b). The predicted T- and B-cell regions show very similar distance distributions that are lower than those in the hilum, and slightly higher than the average distance within the whole network. This trend is also evident when the mean distances within subregions are compared (Fig. 7c). The equivalence of vessel distances in T and B cell regions initially seems counter-intuitive given the low density of vessels within the follicle centre. However, the analysis shown in Fig. 7 not only incorporates the follicular core but also the vasculature immediately surrounding follicles, which compensates for the high vessel distances that can be found within the centre of the follicles (Supplementary Figure 4). We also identified regions of dense vasculature outside these three defined zones, characterised by shorter distances to supplying blood vessels (Fig. 7a–c). To compare our 3D measurements with the conventional method for evaluating tissue sections, we created a 2D histology tool to evaluate vascular density in different compartments in 2D. This tool confirmed large density differences between subcompartments, and measurements of vascular density in 2D had the same hierarchy as in 3D (Fig. 7d). A combination of several different measurements that are feasible within the 3D model provides a new vision of the 3D topology of vessels within specific regions of the LN (Fig. 7f). Although B-cell follicles can appear nearly avascular in 2D images (Fig. 7f, upper panel), 3D visualisation shows the surrounding vascular loops supplying the follicle, as well as a central area within the follicle that is >60 μm distal from any of these vessels, which could be found in the majority of follicles in this study. Nearby HEVs that supply lymphocytes to the follicle can also be highlighted within the same image projections to provide a complete 3D view of the vascular environs (Fig. 7f) that is not evident in 2D views.

Bottom Line: By focussing on critical immune microenvironments we quantified differences in their vascular topology.We further developed a morphology-based approach to identify High Endothelial Venules, key sites for lymphocyte extravasation.These data represent a comprehensive and continuous blood vessel network of an entire organ and provide benchmark measurements that will inform modelling of blood vessel networks as well as enable comparison of vascular topology in different organs.

View Article: PubMed Central - PubMed

Affiliation: Maurice Wilkins Centre, University of Auckland, Private Bag 92-019, Auckland 1142, New Zealand.

ABSTRACT
Understanding of the microvasculature has previously been limited by the lack of methods capable of capturing and modelling complete vascular networks. We used novel imaging and computational techniques to establish the topology of the entire blood vessel network of a murine lymph node, combining 63,706 confocal images at 2 μm pixel resolution to cover a volume of 3.88 mm(3). Detailed measurements including the distribution of vessel diameters, branch counts, and identification of voids were subsequently re-visualised in 3D revealing regional specialisation within the network. By focussing on critical immune microenvironments we quantified differences in their vascular topology. We further developed a morphology-based approach to identify High Endothelial Venules, key sites for lymphocyte extravasation. These data represent a comprehensive and continuous blood vessel network of an entire organ and provide benchmark measurements that will inform modelling of blood vessel networks as well as enable comparison of vascular topology in different organs.

No MeSH data available.


Related in: MedlinePlus