The blood brain barrier (BBB) in the central nervous system (CNS) regulates the passage of nutrients, essential components and metabolic constituents between the blood stream and the parenchymal tissue. Although local neural activity is known to promote local blood supply and is the basis of the so-called blood-oxygen level gradient/dependency (BOLD) 1 2 3 , the regulation of local blood flow in the parenchyma of the CNS is not well understood. Contraction of pre-capillary arterioles has been suggested but so far without clear demonstration 4 . Capillaries of the CNS do not have smooth musculature but contractility of pericytes associated with the capillaries of peripheral tissue such as the retina, cardiac and skeletal muscles has been observed 5 6 7 8 9 , suggesting that contraction of these cells in the CNS can be a mechanism of blood supply regulation 10 11 12 . However, the finding that pericytes of the CNS lack the α-actin protein isoform found in contractile cells 13 , appears to exclude their participation in CNS capillary contraction. Angiogenesis described during brain development 14 has also been suggested as a mechanism of blood flow increment, although without clear support in normal adulthood or non-pathological contexts 15 16 17 18 . The endothelial cells (EC), pericytes and basal lamina (anatomical constituents of the BBB), together with elements of the neurovascular compartment such as neurons, astrocytes, and other glial cells, adapt themselves to maintain homeostasis that promotes tissue survival 19 20 . Although the endothelial cell can be considered as an element which plays an important role in blood flow regulation its participation has been little analyzed, at least partly due to the intrinsic difficulty of its individual staining in capillaries of the CNS 21 22 23 .

However, in order to better understand the mechanisms that occur during local blood flow changes in the CNS parenchyma, the role of capillary morphology needs to be considered as an important issue. Electron microscopy (EM) has provided knowledge of the capillary wall constituents 24 25 , but the small spatial range of this method is a critical and important limitation of this powerful tool, e.g. the size of a capillary endothelial cell is still not known. Despite attempts to obtain markers enabling CNS capillary imaging or visualization of capillary networks through extracellular tracers, fluorescent markers, diamine benzidine reaction, immunocytochemistry as well as brain arterial injection with black ink or venous plastic infusion 1 15 22 23 24 26 , a complete view of the morphology of CNS capillaries still is lacking. In this respect, recently we proposed an alternative histological tool focused on detecting the relative space occupied by the capillary endothelial cell 27 28 . This visualization through light microscopy is based on two facts: first, the optical effect that reflecting light produces in fixed and unstained capillaries, and second, evidence provided by EM 24 25 showing that the internal wall of CNS capillaries is exclusively constituted by the endothelial cell. Using this procedure we found that changes in capillary luminal area can be detected indirectly via changes in the inner capillary diameter (ICD).

With the aim of investigating the participation of the capillary endothelial cell in the regulation of blood flow fluctuations that occur in response to changes in neuronal activity, the present study was directed to quantifying the internal diameter of capillaries in the anterior hypothalamic paraventricular nucleus (PVN) in female rats. We chose this nucleus since several of its physiological aspects but not its microvasculature have been intensely studied 29 30 31 32 33 34 35 . For example, it is well known that the magno- and parvocellular regions via a neuroendocrine reflex are involved in oxytocin (OT) production and release in response to mammary gland sensory stimulation 36 37 , and that the effect varies according to the reproductive phase of the female rat 33 34 35 37 . Thus, we decided to measure and compare the capillary internal diameter during the diestrous phase of virgin rats and during the lactation period of mother rats. To ensure that the sensory information received by the two groups was significantly different, mammary glands of the mother rats were stimulated by their pups suckling for two weeks of lactation, while virgin rats had no stimulation.

Whereas the physiology and neuronal characteristics of the PVN have been widely investigated 29 30 31 32 33 34 35 , its microvasculature has been much less so 29 39 40 . Morphological changes such as glial retraction, neuronal remodeling, c-Fos expression and electrophysiological activity increment during lactation have been reported 35 41 42 43 , and there is no doubt that greater synthesis and release of OT from both the PVN and supraoptic nucleus occurs 33 34 35 . This could reflect an increment in the local neural activity of these structures that in turn promotes an increment in local blood flow 1 2 3 4 . In the present study our results suggest that, at least in the PVN, local blood flow increment could be via an increment in the ICD rather by an increase in outer capillary diameter or angiogenesis. Our findings showing that the area occupied by the neural and glial elements of the neurovascular compartment did not change in virgin compared to lactating rats, suggest that angiogenesis is not present. In addition, it has been suggested that components of the capillary basal lamina such as laminin, cablin, fibronectin and type IV collagen among others 44 45 46 47 prevent the outer EC diameter from expanding.

In relation to our proposal that ICD increment promotes increases in blood flow, we mention the following: Local blood flow regulation in the parenchyma of the CNS is not well understood 4 5 6 7 10 11 12 . However, capillary contraction caused by swelling of EC with no change in the external diameter of the capillary has been previously suggested as an active mechanism of local capillary constriction in peripheral tissue 48 . Thus, in accordance with our findings of significant changes in the ICD of PVN microvasculature in the LAC group, we take up this idea again to offer an account of how fast, regional, and reversible control of blood supply in the CNS might occur. Changes in the ICD might take place via modifications of EC cytoplasmic volume; when this decreases the capillary lumen increases (Figure 7A), and the time needed for cytoplasmic transport of substances from the luminal to the abluminal side of the membrane decreases. It is known that the reduction or increase in either local blood supply or the bidirectional “blood ↔ parenchyma” transport of metabolic constituents depends upon the EC which, together with the basal lamina, constitute the capillary wall 25 47 . Accordingly, when the cytoplasmic volume increases, the capillary lumen decreases (Figure 7B), causing cytoplasmic transport of metabolites from the luminal to the abluminal side of the membrane to be slowed. There are a wide variety of factors that modify intracellular osmolarity, thereby promoting changes in cytoplasmic EC volume. Organic osmolytes, taurine, cyclosporine, water, some hormones such as vasopressin (VP) or OT, and free radicals have all been implicated in mechanisms of cell volume regulation 49 50 51 52 53 . Thus, it is possible that neurons and/or astrocytes could modulate their own capillary blood supply as well as the transport and supply of metabolically important substances via such factors.

In addition, it is currently thought that transport of water-soluble molecules across EC is accomplished by the continuous formation of plasmalemma vesicles followed by detachment and fusion to the membrane on the other side of the cell or by transport of cytoplasmic vesicles moving from one surface to the other without membrane fusion 47 54 55 . Either of these mechanisms of intracellular transport could be present in the capillary EC and depending on the luminal-abluminal distance the cytoplasmic transport time could be faster or slower. This could represent a dynamic and plastic mechanism by which constantly changing neuronal needs are rapidly met. As physiological changes associated with chronic dehydration, hemorrhage, stress condition as well as pregnancy and parturition 33 56 , are associated with the synthesis and release of OT or VP from the PVN and supraoptic nucleus, then, based on our results, we should not reject the idea that a similar dynamic mechanism of blood flow regulation in response to such stimuli might occur.

Finally, it is well-known that during lactation progesterone (P4) and prolactin (PRL) reach high levels in the blood stream 33 . Since PRL has been implicated in OT release 57 58 , this hormone could be a potential candidate to participate in PVN blood flow regulation.

Our findings suggest that during lactation the microvasculature of the PVN of female rats undergoes dynamic, transitory changes in blood supply represented by an increment in the ICD through self-cytoplasmic volume modification reflected by EC changes. This is relevant to consider from the pharmacological point of view, since changes in cytoplasmic volume in the EC could help to increase blood flow and in this way facilitate the metabolic exchange between capillaries and the CNS parenchyma.

BBB: Blood brain barrier; CNS: Central nervous system; EC: Endothelial cell; EM: Electron microscopy; ICD: Inner capillary diameter; LAC: Lactating group; OT: Oxytocin; PRL: Prolactin; PVN: Hypothalamic paraventricular nucleus; VIR: Virgin group; VP: Vasopressin.

The authors declare that they have no competing interests.

ACS participated in the design of the study, performed the experiment, processed and analyzed the data, and drafted the manuscript. MLG provided assistance with data processing and analyses. MA provided recommendations for the experimental design and analyses. RH and PB provided insight, recommendations and edited the manuscript. PP conceived the study, supervised the experimental design and the coordination of all parties involved with the study, and edited the manuscript. All authors read and approved the final draft.