An electron microscope study of the differentiating capillaries of the mouse neurohypophysis

Historical and current perspectives on blood endothelial cell heterogeneity in the brain

Dynamic brain activity requires timely communications between the brain parenchyma and circulating blood. Brain-blood communication is facilitated by intricate networks of brain vasculature, which display striking heterogeneity in structure and function. This vascular cell heterogeneity in the brain is fundamental to mediating diverse brain functions and has long been recognized. However, the molecular basis of this biological phenomenon has only recently begun to be elucidated. Over the past century, various animal species and in vitro systems have contributed to the accumulation of our fundamental and phylogenetic knowledge about brain vasculature, collectively advancing this research field. Historically, dye tracer and microscopic observations have provided valuable insights into the anatomical and functional properties of vasculature across the brain, and these techniques remain an important approach. Additionally, recent advances in molecular genetics and omics technologies have revealed significant molecular heterogeneity within brain endothelial and perivascular cell types. The combination of these conventional and modern approaches has enabled us to identify phenotypic differences between healthy and abnormal conditions at the single-cell level. Accordingly, our understanding of brain vascular cell states during physiological, pathological, and aging processes has rapidly expanded. In this review, we summarize major historical advances and current knowledge on blood endothelial cell heterogeneity in the brain, and discuss important unsolved questions in the field.

The hypothalamic neuropeptide oxytocin is required for formation of the neurovascular interface of the pituitary

The hypothalamo-neurohypophyseal system (HNS) is the neurovascular structure through which the hypothalamic neuropeptides oxytocin and arginine-vasopressin exit the brain into the bloodstream, where they go on to affect peripheral physiology. Here, we investigate the molecular cues that regulate the neurovascular contact between hypothalamic axons and neurohypophyseal capillaries of the zebrafish. We developed a transgenic system in which both hypothalamic axons and neurohypophyseal vasculature can be analyzed in vivo. We identified the cellular organization of the zebrafish HNS as well as the dynamic processes that contribute to formation of the HNS neurovascular interface. We show that formation of this interface is regulated during development by local release of oxytocin, which affects endothelial morphogenesis. This cell communication process is essential for the establishment of a tight axovasal interface between the neurons and blood vessels of the HNS. We present a unique example of axons affecting endothelial morphogenesis through secretion of a neuropeptide.

Physiologic upper limits of pore size of different blood capillary types and another perspective on the dual pore theory of microvascular permeability

BACKGROUND

Much of our current understanding of microvascular permeability is based on the findings of classic experimental studies of blood capillary permeability to various-sized lipid-insoluble endogenous and non-endogenous macromolecules. According to the classic small pore theory of microvascular permeability, which was formulated on the basis of the findings of studies on the transcapillary flow rates of various-sized systemically or regionally perfused endogenous macromolecules, transcapillary exchange across the capillary wall takes place through a single population of small pores that are approximately 6 nm in diameter; whereas, according to the dual pore theory of microvascular permeability, which was formulated on the basis of the findings of studies on the accumulation of various-sized systemically or regionally perfused non-endogenous macromolecules in the locoregional tissue lymphatic drainages, transcapillary exchange across the capillary wall also takes place through a separate population of large pores, or capillary leaks, that are between 24 and 60 nm in diameter. The classification of blood capillary types on the basis of differences in the physiologic upper limits of pore size to transvascular flow highlights the differences in the transcapillary exchange routes for the transvascular transport of endogenous and non-endogenous macromolecules across the capillary walls of different blood capillary types.

METHODS

The findings and published data of studies on capillary wall ultrastructure and capillary microvascular permeability to lipid-insoluble endogenous and non-endogenous molecules from the 1950s to date were reviewed. In this study, the blood capillary types in different tissues and organs were classified on the basis of the physiologic upper limits of pore size to the transvascular flow of lipid-insoluble molecules. Blood capillaries were classified as non-sinusoidal or sinusoidal on the basis of capillary wall basement membrane layer continuity or lack thereof. Non-sinusoidal blood capillaries were further sub-classified as non-fenestrated or fenestrated based on the absence or presence of endothelial cells with fenestrations. The sinusoidal blood capillaries of the liver, myeloid (red) bone marrow, and spleen were sub-classified as reticuloendothelial or non-reticuloendothelial based on the phago-endocytic capacity of the endothelial cells.

RESULTS

The physiologic upper limit of pore size for transvascular flow across capillary walls of non-sinusoidal non-fenestrated blood capillaries is less than 1 nm for those with interendothelial cell clefts lined with zona occludens junctions (i.e. brain and spinal cord), and approximately 5 nm for those with clefts lined with macula occludens junctions (i.e. skeletal muscle). The physiologic upper limit of pore size for transvascular flow across the capillary walls of non-sinusoidal fenestrated blood capillaries with diaphragmed fenestrae ranges between 6 and 12 nm (i.e. exocrine and endocrine glands); whereas, the physiologic upper limit of pore size for transvascular flow across the capillary walls of non-sinusoidal fenestrated capillaries with open 'non-diaphragmed' fenestrae is approximately 15 nm (kidney glomerulus). In the case of the sinusoidal reticuloendothelial blood capillaries of myeloid bone marrow, the transvascular transport of non-endogenous macromolecules larger than 5 nm into the bone marrow interstitial space takes place via reticuloendothelial cell-mediated phago-endocytosis and transvascular release, which is the case for systemic bone marrow imaging agents as large as 60 nm in diameter.

CONCLUSIONS

The physiologic upper limit of pore size in the capillary walls of most non-sinusoidal blood capillaries to the transcapillary passage of lipid-insoluble endogenous and non-endogenous macromolecules ranges between 5 and 12 nm. Therefore, macromolecules larger than the physiologic upper limits of pore size in the non-sinusoidal blood capillary types generally do not accumulate within the respective tissue interstitial spaces and their lymphatic drainages. In the case of reticuloendothelial sinusoidal blood capillaries of myeloid bone marrow, however, non-endogenous macromolecules as large as 60 nm in diameter can distribute into the bone marrow interstitial space via the phago-endocytic route, and then subsequently accumulate in the locoregional lymphatic drainages of tissues following absorption into the lymphatic drainage of periosteal fibrous tissues, which is the lymphatic drainage of myeloid bone marrow. When the ultrastructural basis for transcapillary exchange across the capillary walls of different capillary types is viewed in this light, it becomes evident that the physiologic evidence for the existence of aqueous large pores ranging between 24 and 60 nm in diameter in the capillary walls of blood capillaries, is circumstantial, at best.

Changes in the distribution and fine structure of the intralobular blood vessels of the submandibular gland in the postnatally developing mouse

Previous studies have shown that the blood vessels supplying the endocrine organs and the mucosa of the intestinal canals change in terms of not only their distribution but also their structure with the development and growth of each organ. We examined changes in the distribution and structure of intralobular blood vessels, including capillaries, throughout the postnatal development of the submandibular gland, an exocrine organ. The mouse submandibular gland from days 0 (birth) to 49 was investigated chronologically and ultrastructurally. The capillaries changed from continuous to fenestrated on day 10, coincident with an increase in the number of acini to more than the number of terminal tubules. The number of sections of intralobular blood vessels per unit area gradually decreased with increasing acinar size and was lowest on day 21 when pups were weaned; the same number was maintained from then on. In contrast with the reduction in the number of intralobular blood vessels, the number of capillary pores appeared to increase gradually. Acinar size increased further till day 28. Capillary pore number also increased further, till day 35, apparently in relation to the increasing acinar size. These findings suggest that the changes in distribution and structure of the intralobular blood vessels in the submandibular gland of the postnatally developing mouse are closely related to the development of the parenchymal cells in preparation for weaning and sexual maturity.

The fine structure of the developing human choriocapillaris during the first trimester

In human embryos with a gestational age ranging from 6.5 to 13 weeks the development of the choriocapillaris was studied by electron microscopy. A continuous maturation process was observed over the whole period, which was most pronounced between weeks 6.5 and 9.7. During the earlier weeks, the endothelium appeared to be rather thick and contained many cytoplasmic vesicles; it was markedly flattened and less vesiculated than at week 9.7. This process was accompanied by an enlargement in the vascular lumina. Fenestrations directed towards the primitive Bruch's layer were found to be already present at week 7.2 but increased in number from week 9.7 to week 13. In the early weeks, basement membrane structures were present in small, discontinuous tufts of cloudy material adjacent to the endothelium. These structures became continuous at week 9.7 and increased in thickness towards week 13. In the earlier weeks, punctate junctions were seen. Beginning with week 9.7, adultlike tight junctions between the endothelial cells became observable. Pericytes were already present at week 6.5. It is suggested that both pericytes and endothelial cells are derived from the differentiating mesenchymal cells of the surrounding ocular stroma.

Fine structure of pituitary blood vessels in embryos of the dwarf (dw) mutant mouse

Blood vessels in the pars distalis of the embryonic pituitary gland of the dwarf (dw) mutant mouse were analyzed by means of electron microscopy. Dwarf (dw/dw) embryos ranging in age from 15 through 18 days of gestation were obtained from matings of dwarf homozygotes in which reproductivity was induced by means of hormone supplementation and grafts of normal pituitary to the kidney capsule. Endothelial cells in vessels of the dwarf pituitary were retarded in development as indicated by a retention of macrovesicles and marginal flaps beyond embryonic stages when they normally subside in normal embryos, as well as a lag in the development of endothelial cell attenuations. This retardation may be causally related to the pituitary pathology involving faulty cytodifferentiation of somatotrophs and mammotrophs.

Permeability of the blood-brain barrier in the median eminence during the perinatal period in rats

An electron microscopic study has been carried out in order to examine the permeability of the blood-brain barrier in the median eminence of perinatal rats. After several minutes, intravascularly injected electron-dense tracers (lanthanum nitrate; horseradish peroxidase, 40000 MW1, ferritin, 500000 MW) pass the capillary wall, the perivascular space, and become incorporated into neurosecretory axons and basal processes of tanycytes both in fetuses and young rats. In the case of immature capillaries, the materials diffuse freely through the endothelial cells, and to a lesser extent are transferred via occasional plasmalemmal vesicles and fenestrae. As the maturation of capillaries proceeds, their permeability via plasmalemmal vesicles and fenestrae increases considerably due to a gradual rise of the number of these structures. The plasmalemmata of the differentiated endothelial cells become impermeable to all of the tracers. Only ionic lanthanum appears to penetrate through transendothelial channels and intercellular junctions between adjacent endothelial cells.

An electron microscopic study on the structural development of the neural lobe in the human fetus

Structural development of the neural lobe of the hypophysis was studied by light and electron microscopy in five human fetuses ranging from 7.5 to 19 weeks of ovulation age. Clusters of small clear vesicles were found occasionally in axon profiles in a fetus of 7.5 weeks. Granular vesicles in axon profiles first accumulated in a fetus of 8.5 weeks. Concurrently, fenestration of capillary walls and expansion of perivascular spaces were observed. After 8.5 weeks, three types of vesicle-containing axon profiles were found, which contained predominantly granular vesicles (GP), small clear vesicles (SP), and a mixed population of both granular and small clear vesicles (MX), respectively. Their numbers increased significantly during two periods, 7.5-8.5 weeks and 15.5-19 weeks. Ratios of GP, SP, and MX to the total number of the three types of axon were almost the same in all the fetuses studied, i.e., about 40%, 40%, and 20%, respectively. Developmental changes in the diameter of granular vesicles were examined in GP, SP, and MX of each fetus. Apart from a few exceptions, the mean diameter of granular vesicles was larger in GP than in MX, which was in turn larger than in SP. The diameters increased with development and those in GP reached almost the same value as in the adult by 19 weeks.