Heterogeneity of the afferent arteriole—correlations between morphology and function

The afferent arteriole (AA) is an important structural and regulatory site of blood pressure maintenance and renal salt and water conservation. The AA is considered a typical resistance vessel with ring-like smooth muscle cells in its wall, covered by a homogenous endothelial layer. It also contains renin granular cells mainly at its distal, juxtaglomerular end [ 1 ]. These cells become myosin-negative, round-shaped cells producing a large number of renin granules in their cytoplasm [ 2 ]. The AA is part of a functional syncytium at the vascular pole of renal corpuscle, called the juxtaglomerular apparatus (JGA), which consists of the afferent and efferent arterioles, the macula densa (MD) cells of the thick ascending limb and the extraglomerular mesangium [ 1 ]. Two major regulatory processes have been traditionally attributed to the AA-JGA: the control of vascular resistance and renin release. Both direct mechanisms (myogenic, neurohumoral) and indirect mechanisms (tubuloglomerular and AA short-loop feedback) are involved in the regulation of AA vascular resistance [ 3 ]. Also, it is accepted that the major portion of plasma renin originates from the renin granular cells of the JGA [ 4 ]. The AA has been widely and extensively studied and several comprehensive reviews are available [ 5–8 ]. The present short review focuses on some selected morphological and functional features of the distal portion of AA. In 1889, Golgi [ 9 ] recognized that the distal portions of AAs were in intimate contact with the distal part of the renal tubule which always returned to its parent glomerulus, and suggested that this close contact should have some regulatory consequences. In 1925, Ruyter [ 10 ] described the existence of JGA granular epithelioid cells and assumed that they were involved in the regulation of renal blood flow (RBF) and glomerular filtration rate (GFR) by their swelling and shrinking. Goormaghtigh [ 11 ] was the first in 1937 to accurately describe the MD and JGA, and outlined their regulatory role in renal haemodynamics. According to his hypothesis, the resistance of the distal part of AA is regulated humorally. The humoral mediator is liberated upon the activity of MD, which senses the ‘state of filling of the interposed nephron segment’ or ‘the physico-chemical composition of the urine that flows by’. In 1957, Hársing et al . [ 12 ] were the first to demonstrate the existence of tubuloglomerular feedback (TGF) in whole animal studies. They stated that after diuretic treatment in anaesthesized dogs, ‘a correlation was found to exist between the diuresis expressed in percentage of the filtrate and the relative reduction in RBF and GFR’. During TGF, MD cells detect changes in distal tubular flow and salt concentration and transmit signals through the extraglomerular mesangium to the smooth muscle cells of AA to regulate GFR. This is called the ‘long-loop feedback mechanism’. MD cells also contribute to the complex regulation of renin release from the epithelioid/granular cells of the JGA which is also controlled by a number of other direct or indirect influences (neuro-hormonal, local baroreceptors) [ 13 ]. TGF studies suggest that the activation of MD may result in two opposing events. The immediate effect would cause vasoconstriction of the AA with the consequence of reduction in GFR, within seconds of the increase in distal tubular salt. The same increase in salt delivery decreases renin secretion, which consequently leads to decreases in angiotensin levels, and ultimately results in vasodilation of AA and an increased GFR. It is postulated that the TGF mechanism and its rapid activation help to avoid large fluctuations in GFR by regulating vascular resistance of the AA, when salt delivery to the distal nephron changes [ 3 ]. Recently, a novel experimental approach has been developed using the isolated in vitro microperfused JGA to study the function of the JGA complex in its total integrity, yet independent from the surrounding tissues [ 14 ]. The multiphoton fluorescence imaging technique allows real-time visualization of both TGF and renin release mechanisms [ 15 ]. Using this technique, dynamics of both the release and tissue activity of renin can be studied at the sub-cellular level with high temporal and spatial resolution ( Figure 1 ). We have recognized that the endothelium of the distal (juxtaglomerular) part of AA is morphologically different from the endothelium covering the proximal portion (i.e. upstream, towards the interlobular artery). We have demonstrated endothelial fenestrations in the juxtaglomerular AA in different mammals including humans [ 16 , 17 ]. These fenestrae face the extraglomerular mesangium and renin granular epithelioid cells ( Figure 2 ). The pores have uneven diameters ranging between 60 and 240 nm. The structure of these fenestrae is similar to those in the intra-glomerular endothelium, in contrast to the endothelial fenestration described in other organs. Namely, there is no glycoprotein diaphragm spanning the pores. This fenestration allows bulk fluid flow from the lumen of AA towards the interstitium of JGA and we have demonstrated this by using ferritin particles as an indicator of permeability [ 18 ]. Also, in a preliminary study using multiphoton imaging [ 19 ] the fluid marker lucifer yellow given by i.v. bolus appeared in the juxtaglomerular AA lumen and in the surrounding interstitium before the dye reached the glomerular capillaries. This flow of fluid from lumen of the arteriole or glomerular capillaries into the mesangial field may alter the concentration of humoral mediators in the interstitium of the JGA, which may be critical in the regulation of its function. Using Amphiuma , whose kidney has fenestrations in the distal part of AA and shows an active TGF mechanism, we were able to demonstrate changes in glomerular capillary pressure in response to changes in AA colloid-osmotic pressure [ 13 ]. In the Amphiuma , the efferent (post-glomerular) arteriole does not feed a peritubular capillary network, unlike in the mammalian kidney. Thus, manipulation of the colloid-osmotic pressure of AA has no osmotic effect on the tubular interstitium around the parent nephron. In this way, we were able to demonstrate a vasculo-vascular ‘short-loop’ mechanism regulating glomerular capillary pressure and filtration rate ( Figure 3 ). The existence of this mechanism has not been proven in mammals yet, but based on the presence of the morphological prerequisites we suggest that this mechanism operates in the human kidney as well. It is known that the main source of renin in the circulating blood is the AA granular cells [ 4 ]. It is thought that renin is secreted both into the interstitium and the lumen of the AA. The luminal secretion of the hormone is suggested by the observation that the concentration of renin is higher in the post-glomerular arteriole compared with that of the pre-glomerular arteriole [ 20 ]. In earlier work using serial sections of fixed rat kidney, consecutively immunostained for myosin and renin, it was demonstrated that the renin-positive segment of AA is myosin-negative [ 2 , 21 ]. Furthermore, a significant correlation between the renin-positive and fenestrated (permeable) endothelium portion of AA was found [ 21 ]. It has long been unclear how renin, a large molecule (∼40 kDa) could traverse the endothelial layer covering the renin granular cells. It has been demonstrated that the renin-positive juxtaglomerular portion of AA has fenestrated endothelium facing the renin granular cells in experimental animals including Amphiuma, Tupaia belangeri (tree shrew), rats and humans [ 2 , 16 , 17 , 21 ]. Also, it has been shown in both rat and human kidney that the parietal layer of Bowman's capsule faces the renin-producing portion of AA and is covered by foot processes forming filtration slits [ 2 ]. Based on these findings, we suggest that the endothelial fenestration may provide a communication pathway between renin granular cells and the lumen of AA. Thus, the unique endothelial morphology of the distal AA might be instrumental in the regulation of renin release into the blood. Supporting this hypothesis, we were able to visualize endothelial fenestration adjacent to renin granules in vivo using multiphoton fluorescence imaging ( Figure 4 ). Based on our findings, we have concluded that the AA is not a vessel with uniform characteristics along its entire length. It can be divided into two structurally and functionally different portions: a myosin-positive, renin-negative proximal portion lined by non-permeable endothelium and a myosin-negative, renin-positive distal portion lined by fenestrated, highly permeable endothelium. It has been demonstrated that the length of each segment is highly variable, tracking the actual status and activity of the renin–angiotensin system (RAS) [ 21 ] ( Figure 5 ). At birth, the entire length of AA is renin-positive lined by permeable endothelium, while the length of this segment is progressively reduced with aging (unpublished data). The mechanism of the transition from one type to the other during aging and during changes in the activity of RAS, remains to be elucidated. Fenestrated endothelium without diaphragms can only be found in distinct regions of the circulation such as the glomeruli and endocrine organs [ 22 , 23 ]. The fenestration in the AA endothelium is peculiar since (i) it exists in a blood vessel in which the intravascular pressure is high and (ii) it is uniformly present in various species from amphibians to mammals. However, it is unclear at present how these pores are formed and how their size is controlled. It is suggested that the vascular endothelial growth factor (VEGF), a major regulator of vascular permeability, is one of the key players in this phenomenon:Based on these data, we suggest that VEGF plays a key role in the formation of endothelial fenestration in the juxtaglomerular portion of AA. This hypothesis is supported by our preliminary data i.e. the development of fenestration in endothelial cell culture following VEGF treatment. In other laboratories, expression of VEGF was found to be tightly regulated by several factors that are related to the JGA. For example, angiotensin II, the main effector of the RAS, stimulated VEGF secretion [ 28 ]. This may be the key mechanism in the distal AA as well, and it could explain the high correlation between fenestration and renin expression (RAS activity). Relaxin, a member of the insulin hormone family has been recently recognized as another factor that stimulates VEGF expression in different tissues both in vitro and in vivo [ 29 , 30 ]. Interestingly, relaxin was shown to stimulate renin secretion in decidual cells of the placenta. Plasma relaxin levels are known to be high in the last trimester of pregnancy [ 31 ]. This is consistent with our finding on increased renin expression and endothelial fenestration in the AA during the early days of life ( Figure 6 ). It was demonstrated in vivo that the fenestrated endothelium is always associated with an epithelial layer strongly expressing VEGF, and it was postulated that VEGF is the factor maintaining endothelial fenestration [ 24 ]. Although many factors are involved in the regulation of vascular permeability, most of them exert their effects only on intercellular junctions and mainly in post-capillary venules. In contrast, VEGF was shown to regulate fenestration in many different vessels [ 22 ]. Although the expression of VEGF in adults is low, it is relatively high in the kidney. Marked expression was detected in glomerular podocytes, in distal tubular epithelial cells and in the muscular layer of arterioles [ 25 ]. VEGF receptors are present on renal endothelial and tubular cells [ 26 , 27 ]. The renal AA consists of two morphologically and functionally distinct segments: the contractile proximal part upstream and the renin-rich distal, juxtaglomerular portion lined with a highly permeable endothelium. The distal, endocrine-like portion may not only participate in renin secretion from the JGA, but also in the regulation of the fluid balance in the interstitium of the JGA. By this latter function the fenestrated portion of AA is directly involved in the regulation of glomerular capillary pressure and filtration. The length of this portion is variable and controlled by several mechanisms that need further study. Visualization of the in vitro microperfused afferent arteriole (AA) with attached glomerulus (G) and macula densa (MD). Compared with the proximal AA segment, the distal, juxtaglomerular part contains renin producing granular cells. Differential interference contrast (DIC) image ( A ) and confocal fluorescence image ( B ) of the same preparation using the cell membrane marker R-18 (AA endothelium, red) and quinacrine (green), a fluorophore selective for renin granules. Different endothelial morphology in the renin-negative and positive AA segments is indicated by the intense and continuous red labelling in the proximal segment compared with the weak and interrupted labelling in the distal part, respectively. Bar is 20 µm. (Unpublished data.) Fenestrated endothelium in the distal portion of the afferent arteriole as shown in the inset in rat. MD, macula densa; LEA, lumen of the efferent arteriole; LAA, lumen of the afferent arteriole; I, interstitial space; BC, Bowman capsule; GST, glomerular stalk; U, urinary space; (electron micrograph). (From [ 13 ].) The scheme of the classical ‘long-loop’ and the ‘short-loop’ mechanism playing a role in the regulation of GFR and RBF. VSMC, vascular smooth muscle cells; JGEC, juxtaglomerular epithelioid cells. (From [ 13 ].) Multiphoton image of the distal portion of afferent arteriole (AA). Horizontal ( A ) and cross-sectional ( B ) view of the fenestrated endothelium and AA lumen. The endothelial layer is labelled by the endocytosis of lucifer yellow. Content of individual renin granules is labelled by quinacrine (green). Bar is 5 µm. (Unpublished data.) Schematic characterization of the two different arteriolar segments. These segments represent distinct morphological and functional features and their length is variable. (From [ 13 ].) The role of VEGF, RAS and relaxin in the regulation of endothelial permeability. (Dashed arrow: inhibitory effect.) The studies were supported by Hungarian Research Grants: OTKA AT 048767, ETT 564/2003, Hungarian Kidney Foundation (to L.R.) and NIH DK064324 (to J.P.-P). Conflict of interest statement . None declared.