Endothelial cells modulate renin secretion from isolated mouse juxtaglomerular cells.

Utilizing cocultures of mouse renal juxtaglomerular cells with bovine microvascular endothelial cells, we have examined whether endothelial cells exert direct influence on renin secretion from renal juxtaglomerular cells. In the presence of endothelial cells both spontaneous and forskolin (10 microM) or isoproterenol (10 microM) stimulated renin release were markedly attenuated. The stimulatory effect of the calmodulin antagonist calmidazolium (10 microM) on renin secretion was not altered by endothelial cells, whereas the stimulatory effect of ethylisopropylamiloride (50 microM) an inhibitor of sodium-proton exchange was enhanced in the presence of endothelial cells. Indomethacin (10 microM) and NG-monomethyl-l-arginine (NMMA) (1 mM) used to inhibit cyclooxygenase activity and production of endothelium-derived relaxing factor (EDRF) decreased spontaneous renin release in the presence of endothelial cells only, but had no effect on forskolin stimulated renin secretion. Endothelin (1 microM) inhibited cAMP stimulated renin release both in the absence and in the presence of endothelial cells. ATP (10 microM) which acts on both endothelial and juxtaglomerular cells via purinergic P2 receptors inhibited cAMP stimulated renin release only in the absence but not in the presence of endothelial cells. This modulatory effect of endothelial cells was no altered by indomethacin nor by NMMA. Taken together, our findings provide first evidence for a local control function of the endothelium on cAMP stimulated renin secretion from renal juxtaglomerular cells, which could in part be mediated by endothelin.

Role of microtubules in the rapid regulation of renal phosphate transport in response to acute alterations in dietary phosphate content.

Renal proximal tubular response to acute administration of a low Pi diet is characterized by a rapid adaptive increase in apical brush border membrane (BBM) Na-Pi cotransport activity and Na-Pi cotransporter protein abundance, independent of a change in Na-Pi cotransporter mRNA levels (Levi, M., M. Lötscher, V. Sorribas, M. Custer, M. Arar, B. Kaissling, H. Murer, and J. Biber. 1994. Am. J. Physiol. 267: F900-F908). The purposes of the present study were to determine if the acute adaptive response occurs independent of de novo protein synthesis, and if microtubules play a role in the rapid upregulation of the Na-Pi cotransporters at the apical BBM. We found that inhibition of transcription by actinomycin D and translation by cycloheximide did not prevent the rapid adaptive response. In addition, in spite of a 3.3-fold increase in apical BBM Na-Pi cotransporter protein abundance, there was no change in cortical homogenate Na-Pi cotransporter protein abundance. Pretreatment with colchicine, which resulted in almost complete disruption of the microtubular network, abolished the adaptive increases in BBM Na-Pi cotransport activity and Na-Pi cotransporter protein abundance. In contrast, colchicine had no effect on the rapid downregulation of Na-Pi cotransport in response to acute administration of a high Pi diet. We conclude that the rapid adaptive increase in renal proximal tubular apical BBM Na-Pi cotransport activity and Na-Pi cotransporter abundance is independent of de novo protein synthesis, and is mediated by microtubule-dependent translocation of presynthesized Na-Pi cotransporter protein to the apical BBM.

Rapid downregulation of rat renal Na/P(i) cotransporter in response to parathyroid hormone involves microtubule rearrangement.

Renal proximal tubule cells express in their apical brush border membrane (BBM) a Na/P(i) cotransporter type IIa that is rapidly downregulated in response to parathyroid hormone (PTH). We used the rat renal Na/P(i) cotransporter type IIa (NaPi-2) as an in vivo model to assess early cellular events in the rapid downregulation of this transporter. When rats were treated with PTH for 15 minutes, NaPi-2 abundance in the BBM was decreased. In parallel, transporter accumulated in intracellular vesicles. Concomitantly, microtubules (MTs) were found to form dense bundles of apical-to-basal orientation. After 60 minutes of PTH action, the cells were vastly depleted of NaPi-2, whereas their microtubular cytoskeleton had returned to its normal appearance. Prevention of MT rearrangement by taxol resulted in accumulation of NaPi-2 in the subapical cell portion after 15 minutes and a strong delay in depletion of intracellular transporter after 60 minutes of PTH action. Furthermore, the subapical accumulation of NaPi-2 was associated with the expansion of dense apical tubules of the subapical endocytic apparatus (SEA). Depolymerization of MTs by colchicine likewise caused a retardation of intracellular NaPi-2 depletion. These results suggest that NaPi-2 is downregulated in response to PTH through a rapid endocytic process in 2 separate steps: (a) internalization of the transporter into the SEA, and (b) its delivery to degradative organelles by a trafficking mechanism whose efficiency depends on a taxol-sensitive rearrangement of MTs.

Beta-adrenergic stimulation of renin expression in vivo.

OBJECTIVE: To characterize the effect of beta-adrenergic activation on renal production of renin in the intact organism. DESIGN AND METHODS: Renal expression of renin [renin messenger RNA (mRNA)], plasma renin activity (PRA), inactive renin level, intrarenal renin distribution (immunohistochemistry), and the time course of activation of renin as well as hemodynamic parameters were determined during the subcutaneous infusion of isoproterenol (ISO) into rats. To examine whether beta-adrenergic activation of the renin system is modulated by the rate of salt intake rats were fed diets with normal, low and high salt contents. RESULTS: Systolic blood pressure was not altered. PRA was elevated as much as fourfold after 40 h of ISO infusion. Although renal renin mRNA levels were elevated dose-dependently up to 4.2-fold, no significant recruitment of renin-containing cells could be detected. The time course of PRA revealed a marked transient rise of PRA during 6 h of ISO infusion with a subsequent decline. Inactive renin level was elevated during 3 to 18 h of ISO infusion. In contrast, renin mRNA level increased steadily with a lag phase of 3 h. Infusion of ISO increased PRA and renin mRNA level under a high-salt diet, but had no additional effect either on PRA or on renin mRNA level under low-salt diet. CONCLUSION: Activation of beta-adrenergic receptors is a powerful stimulus of renin secretion and renin gene expression in juxtaglomerular cells in vivo, albeit the kinetics of upregulation of renin secretion and renin expression are markedly different. Therefore, the sympathetic tone might be a major factor determining the activity of the renin system in vivo. The ability of adrenergic agonists to stimulate the renin system appears to be modulated by the steady-state level of salt intake.

Ecto-5'-nucleotidase: localization in rat kidney by light microscopic histochemical and immunohistochemical methods.

We demonstrated the distribution pattern of ecto-5'-nucleotidase (5'-Nu) in rat kidney by enzymatic activity (lead salt precipitation) and by immunohistochemistry with a polyclonal antibody raised in rabbits. Enzyme activity was found in the brush border of the proximal tubule, highest in the P1 segments with decreasing intensity in the P2 segments and weakest in P3 segments in the medullary rays of the cortex. The P3 segments of the outer stripe showed slightly higher activity. Activity was also apparent in the intercalated cells in the connecting tubule and collecting duct, whereas all other tubular and glomerular structures were negative. Activity in peritubular and perivascular connective tissue was highest in the cortical labyrinth, weak or absent in the medullary rays of the cortex, and entirely absent in the medulla. The distribution of the antigen was fully congruent with that of the enzyme activity. With respect to the role of adenosine in regulation of renal blood flow and glomerular filtration rate, the distribution of 5'-Nu in the cortical interstitium may be particularly significant. The possibility of nucleotide cleavage at the brush-border membranes may be important for salvage of nucleotides from the tubular lumen.

Effect of amlodipine on renin secretion and renin gene expression in rats.

1. This study was done to characterize the influence of calcium channel blockade on renin secretion and renin gene expression in normal rats and rats with renovascular hypertension. To this end we studied the effects of the 1,4-dihydropyridine derivative, amlodipine, on plasma renin activity and renal renin m-RNA levels in normal rats and rats with unilateral renal hypoperfusion induced by applying 0.2 mm left renal artery clips over four days. 2. In normotensive rats, amlodipine significantly decreased basal blood pressure by about 20 mmHg when applied in a concentration of 5, 15 and 45 mg kg-1. Plasma renin activity and also renin mRNA levels were not changed after application of 5 mg kg-1 of amlodipine. However, at a concentration of 15 or 45 mg kg-1, amlodipine, significantly increased not only plasma renin activity by about 250% and 300%, but also renin mRNA levels by about 100% and 500%. The action of amlodipine on all these parameters was maximal after 24 h. Treatment with amlodipine in a concentration of 15 mg kg-1 also increased renin immunoreactive areas in the kidney cortex by retrograde recruitment of renin expressing cells in the afferent arterioles. 3. In 2kidney-1 clip rats, systolic blood pressure rose continuously whilst plasma renin activity and renin m-RNA in the clipped kidney increased transiently and renin m-RNA in the contralateral kidney was constantly suppressed. Amlodipine at a concentration of 15 mg kg-1 markedly attenuated the increase of blood pressure in 2kidney-1 clip rats, produced an almost additive effect on plasma renin activity and showed a tendency to increase renin m-RNA levels in the clipped kidneys. Renin m-RNA levels in the contralateral kidney were also significantly suppressed in the animals receiving additional treatment with amlodipine. 4. These findings suggest that inhibition of calcium channels by amlodipine stimulates renin secretion and renin gene expression in vivo. These stimulatory effects are almost additive to the changes of renin secretion occurring after an unilateral fall of renal perfusion pressure.

Enhanced osteopontin expression and macrophage infiltration in MRL-Fas(lpr) mice with lupus nephritis.

MRL-Fas(lpr) mice spontaneously develop a chronic lupus-like renal disease, characterized by immune complex-mediated glomerulonephritis and abundant mononuclear cell infiltration in the interstitium. In the present study we have examined whether the macrophage chemoattractant osteopontin (Opn) could be important in the recruitment of macrophages in this murine model of autoimmune renal injury. We have examined the expression of Opn in the kidney of MRL-Fas(lpr) mice and have correlated Opn synthesis with the degree of macrophage infiltration. Immunofluorescence staining revealed prominent expression of Opn by proximal tubules in MRL-Fas(lpr) mice but not in MRL-++ control mice. Northern blot analysis demonstrated that steady-state transcript levels for Opn mRNA were also significantly increased in MRL-Fas(lpr) kidneys compared with control kidneys. Furthermore, in situ hybridization showed massive Opn mRNA transcripts in proximal tubules in MRL-Fas(lpr) mice but not in controls. The diffuse macrophage infiltration in the kidney of MRL-Fas(lpr) correlated with the enhanced Opn expression. Opn secretion in vitro by cultured renal tubular epithelial cells was upregulated by TNF-alpha and 1,25(OH)2-vitamin D3, whereas no regulation was observed in a control macrophage cell line. We conclude that the enhanced expression of the chemotactic molecule Opn by tubular cells is a prominent feature of murine lupus nephritis and might be promoted by the proinflammatory cytokine environment in MRL-Fas(lpr). The chronic upregulation of Opn could participate in the recruitment of monocytes in the kidney of MRL-Fas(lpr) mice, thereby contributing to the pathogenesis of autoimmune renal disease.

Variability of intercellular spaces between macula densa cells: a transmission electron microscopic study in rabbits and rats.

The macula densa in rabbits and rats was studied by transmission electron microscopy (TEM). After different experimental procedures, the kidneys were fixed by direct perfusion with the fixative without prior flushing of the blood. With this technique, the renal cortex is consistently well preserved. The lateral intercellular spaces of all proximal and distal nephron segments were constantly found to be closed, whereas those of the macula densa varied: depending on the functional situation of the kidney, they were found to be either dilated or closed. Dilated intercellular spaces in the macula densa were encountered in control, sodium-rich, and sodium-deficient rabbits and rats and, in addition, in rats with hypotonic and isotonic hypervolemia. In contrast, in rats with hypertonic hypovolemia and in rats undergoing furosemide or mannitol diuresis, the lateral intercellular spaces of the macula densa were closed. Whether these findings reflect the in vivo state of the macula densa interspaces remains uncertain. The association with specific functional stages demonstrates, at least, a specific behavior of the macula densa cells that is different from that of all other proximal and distal nephron segments and appears to be similar to that of the collecting duct epithelium. The findings suggest that the macula densa is a water-permeable cell plaque within the otherwise water-impermeable thick ascending limb.

Cellular/molecular control of renal Na/Pi-cotransport.

A type II Na/Pi-cotransporter located in the brush border membrane is the rate limiting and physiologically regulated step in proximal tubular phosphate (Pi) reabsorption. In states of altered Pi-reabsorption [for example, in response to parathyroid hormone (PTH) and to altered dietary intake of Pi or as a consequence of genetic abnormalities], brush border expression of the type II Na/Pi-cotransporter is accordingly modified. PTH initiates a regulatory cascade leading to membrane retrieval, followed by lysosomal degradation of this transporter; recovery from inhibition requires its de novo synthesis. Pi-deprivation leads to an increased brush border expression of transporters that does not appear to require de novo synthesis in the short term. Pi-overload leads to membrane retrieval and degradation of transporters. Finally, in animals with genetically altered Pi-handling (Hyp; Gy) the brush border membrane expression of the type II Na/Pi-cotransporter is also reduced, suggesting that a genetically altered protein (such as PEX in Hyp) controls the expression of this transporter.

Three-dimensional shape of a Goormaghtigh cell and its contact with a granular cell in the rabbit kidney.

Goormaghtigh cells of the JGA are characterized by an extensive cellular ramification. In order to elucidate the shape and arrangement of the cell processes a three-dimensional model of a Goormaghtigh cell and of an adjacent granular cell has been constructed based on electron micrographs of a series of ultrathin sections. The model shows that a Goormaghtigh cell has the shape of a flatly pressed cylinder with both ends splitting up into a bunch of parallel processes. The processes maintain a close neighboring position and do not intermingle with processes of other Goormaghtigh cells. This feature is most puzzling when considering that Goormaghtigh cells and their processes are extensively connected by gap junctions. Even processes belonging to the same cell are electrically coupled with each other through gap junctions. The granular cells are clearly different in shape from Goormaghtigh cells. In granular cells bunches of processes are lacking. Granular cells obviously ramify into a few, large processes. The present findings are consistent with the assumption of a functionally central position of Goormaghtigh cells within the feedback mechanism of the JGA.

Morphology of interstitial cells in the healthy kidney.

Renal interstitial cells play an important role in renal function and renal diseases. We describe the morphology of renal interstitial cells in the healthy kidney. We distinguish within the renal interstitium (1) renal fibroblasts and (2) cells of the immune system. Fibroblasts are in the majority and constitute the scaffold of the kidney; they are interconnected by junctions, and are attached to tubules and vessels. Although the phenotype of fibroblasts shows some variation depending on their location in the kidney and on their functional stage, their recognition as fibroblasts is possible on account of structural features. Among the cell types of the second group, antigen-presenting dendritic cells are the most abundant in in the peritubular interstitial spaces of healthy kidneys. Their incidence is highest in the inner stripe of the outer medulla. They share some morphological features with fibroblasts but lack others--junctional complexes, morphologically defined connections with tubules and vessels, and the prominent layer of actin filaments under the plasma membrane--that are characteristic for fibroblasts. Dendritic cells in healthy kidneys are morphologically different from macrophages, which are characterized by abundant primary and secondary lysosomes. In healthy kidneys macrophages are restricted to the connective tissue of the renal capsule and the pelvic wall, and to the periarterial connective tissue. Lymphocytes are rare in healthy kidneys. The distinction of cell types by morphology is supported by differences of membrane proteins. Among all interstitial cells in the renal cortex, fibroblasts alone exhibit ecto-5'-nucleotidase. Dendritic cells constitutively have a high abundance of MHC class II protein. Both proteins are mutually exclusive. Rat macrophages display the membrane antigen ED 2 and lymphocytes exhibit specific surface antigens, depending on their type and functional stage, e.g., CD4 or CD8.

The structural organization of the kidney of the desert rodent Psammomys obesus.

The architecture of the desert rodent Psammomys obesus has been studied by means of standard histologic procedures and by single nephron injections. As other rodent kidneys (rat, mouse), the Psammomys kidney consists of two types of nephrons, 66% short looped and 34% long looped nephrons. The cortex is composed of 4 to 5 layers of glomeruli, which lie closely put together, the glomeruli often touch each other. The superficial and the midcortical glomeruli give rise to short looped neophrons, the juxtamedullary to long looped nephrons. In the strongly developed medulla the inner stripe shows the most striking pattern. It consists of two distinct compartments, that of the giant vascular bundles and that of the interbundle regions. The giant vascular bundles consist of about 8 to 14% arterial vasa recta and 39 to 47% venous vasa recta; furthermore they include the thin descending limbs of the short loops of Henle which amount to 44 to 51% of the bundle structures. The tubules of the interbundle regions surround the bundles in a regular pattern. The inner zone is almost completely surrounded by the renal pelvis; the long broad papilla protrudes into the ureter. The thin descending limbs of short looped nephrons traverse the inner stripe inside the giant vascular bundles. Leaving the bundles they turn back within the inner stripe; their ascending limbs lie in the interbundle region. Both limbs of the long loops of Henle run in the interbundle region, together with the ascending limbs of the short loops and the collecting ducts. The long loops penetrate deeply the inner zone. Many bends are found near the tip of the papilla. The renal pelvis has a very specialized form. It penetrates the inner stripe with many complexely shaped extensions, which surround the giant vascular bundles. Large parts of the bundles with their thin walled structures are thus separated from the pelvic urine only by a single layer of cuboidal epithelium. The possible functional importance of the described specializations of the Psammomys kidney (giant vascular bundles, large inner zone, special shape of the renal pelvis) for the urine concentrating and urea recyclng mechanisms is discussed.

Postnatal maturation of renal cortical peritubular fibroblasts in the rat.

The stromal cells in the renal cortex and medulla of adult rats reveal different phenotypes. Cortical peritubular fibroblasts are ecto-5'nucleotidase (5'NT)-positive and lack alpha-smooth muscle actin (alphaSMA) and vimentin immunoreactivity, whereas medullary fibroblasts are 5'NT-negative and vimentin-positive. We have studied by immunohistochemistry the postnatal (neonatal up to 8 weeks) development of renal cortical stromal cells with respect to 5'NT and to the cytoskeletal proteins alphaSMA and vimentin. Both alphaSMA and vimentin are characteristic for the renal myofibroblasts that replace stromal fibroblasts in interstitial nephritis. In new-born and 1-week-old rats, stromal cells in the cortex and medulla display alphaSMA and vimentin, but lack 5'NT. During the second postnatal week, alphaSMA and vimentin immunoreactivity in cortical interstitial cells gradually declines, whereas 5'NT reactivity becomes progressively apparent between the convoluted tubules in the juxtamedullary labyrinth. For a short time, all three proteins are found to be coexpressed in the same cells. At the end of the third week, interstitial 5'NT-immunoreactivity becomes evident also in the superficial cortical labyrinth, and alphaSMA and vimentin are no longer detectable in cortical peritubular cells. From the fourth week on, the distribution pattern and phenotype of 5'NT-positive cortical fibroblasts correspond to that in adult rats. The temporal pattern of maturation of cortical peritubular fibroblasts seems to parallel the functional maturation of cortical tubules. It is suggested that the local phenotype of peritubular fibroblasts in healthy and possibly also in injured kidneys may be controlled, at least in part, by the local tubular environment, conditioned by tubular metabolism and function.

The role of the BK channel in potassium homeostasis and flow-induced renal potassium excretion.

The kidney is the major regulator of potassium homeostasis. In addition to the ROMK channels, large conductance Ca(2+)-activated K(+) (BK) channels are expressed in the apical membrane of the aldosterone sensitive distal nephron where they could contribute to renal K(+) secretion. We studied flow-induced K(+) secretion in BK channel alpha-subunit knockout (BK(-/-)) mice by acute pharmacologic blockade of vasopressin V(2) receptors, which caused similar diuresis in wild-type and knockout mice. However, wild-type mice, unlike the BK(-/-), had a concomitant increase in urinary K(+) excretion and a significant correlation between urinary flow rate and K(+) excretion. Both genotypes excreted similar urinary amounts of K(+) irrespective of K(+) diet. This was associated, however, with higher plasma aldosterone and stronger expression of ROMK in the apical membrane of the aldosterone-sensitive portions of the distal nephron in the knockout than in the wild-type under control diet and even more so with the high-K(+) diet. High-K(+) intake significantly increased the renal expression of the BK channel in the wild-type mouse. Finally, despite the higher plasma K(+) and aldosterone levels, BK(-/-) mice restrict urinary K(+) excretion when placed on a low-K(+) diet to the same extent as the wild-type. These studies suggest a role of the BK channel alpha-subunit in flow-induced K(+) secretion and in K(+) homeostasis. Higher aldosterone and an upregulation of ROMK may compensate for the absence of functional BK channels.

The renal Na+/phosphate cotransporter NaPi-IIa is internalized via the receptor-mediated endocytic route in response to parathyroid hormone.

The major renal Na(+)/phosphate cotransporter, NaPi-IIa, is regulated by a number of factors including parathyroid hormone (PTH), dopamine, and dietary phosphate intake. PTH induces the acute internalization of NaPi-IIa from the brush border membrane (BBM) and its routing to and subsequent degradation in lysosomes. Previous work indicated that megalin, part of the apical receptor-mediated endocytic apparatus, may play a role in the PTH-induced removal of NaPi-IIa. Here we examined in rats the time-dependent internalization route of NaPi-IIa after acute PTH application using immunohistochemistry and markers of several endocytic compartments. NaPi-IIa removal from the BBM was detectable as early as 5 min after PTH injection. After 10-15 min, NaPi-IIa was localized in subapical compartments positive for clathrin. Shortly thereafter, NaPi-IIa appeared in endosomes stained for EEA1 (early endosomal antigen 1). After 45-60 min, NaPi-IIa was found in late endosomes/lysosomes marked with lgp120. In contrast, no change in the subcellular localization of megalin and the Na(+)/H(+) exchanger NHE3 was detected up to 60 min after PTH injection. To further characterize the internalization route, insulin, as a marker for receptor-mediated endocytosis, and horseradish peroxidase (HRP) and fluorescein isothiocyanate (FITC)-dextran (10 kDa), as markers for fluid-phase mediated endocytosis, were used. NaPi-IIa colocalized with insulin 5-30 min after PTH injection but did not overlap with HRP or FITC-dextran. These results demonstrate a distinct internalization route of NaPi-IIa in response to acute PTH application that may involve the receptor-mediated endocytic pathway including clathrin-coated vesicles and EEA1-positive early endosomes, and routes NaPi-IIa to lysosomes for degradation.

Metabolomic study of cisplatin-induced nephrotoxicity.

We have shown that cisplatin inhibits fatty acid oxidation, and that fibrate treatment ameliorates renal function by preventing the inhibition of fatty acid oxidation and proximal tubule cell death. Urine samples of mice treated with single injection of cisplatin (20 mg/kg body weight) were collected for 3 days and analyzed by 1H-nuclear magnetic resonance (NMR) spectroscopy. In a separate group, urine samples of mice treated with peroxisome proliferator-activated receptor-alpha (PPARalpha) ligand WY were also analyzed by NMR after 2 days of cisplatin exposure. Biochemical analysis of endogenous metabolites was performed in serum, urine, and kidney tissue. Electron microscopic studies were carried out to examine the effects of PPARalpha ligand and cisplatin. Principal component analysis demonstrated the presence of glucose, amino acids, and trichloacetic acid cycle metabolites in the urine after 48 h of cisplatin administration. These metabolic alterations precede changes in serum creatinine. Biochemical studies confirmed the presence of glucosuria, but also demonstrated the accumulation of nonesterified fatty acids, and triglycerides in serum, urine, and kidney tissue, in spite of increased levels of plasma insulin. These metabolic alterations were ameliorated by the use of PPARalpha ligand. Electron microscopic analysis confirmed the protective effect of the fibrate on preventing cisplatin-mediated necrosis of the S3 segment of the proximal tubule. Our study shows that cisplatin-induces a unique NMR metabolic profile in urine of mice that developed acute renal failure, and confirms the protective effect of a fibrate class of PPARalpha ligands. We propose that the injury-induced metabolic profile may be used as a biomarker of cisplatin-induced nephrotoxicity.

Ultrastructural organization of the transition from the distal nephron to the collecting duct in the desert rodent Psammomys obesus.

The transition from the nephron to the collecting duct is formed by three tubular segments (convoluted part of the distal tubule, connecting tubule, cortical collecting duct), which in the desert rodent, Psammomys obesus, transform gradually from one segment to the next, due to intermingling of their different cell types. The convoluted part of the distal tubule (DTC) starts abruptly, shortly beyond the macula densa and initially is homogeneously composed of characteristic DTC-cells. Subsequently, the DTC-cells intermingle with intercalated cells. The first appearance of the connecting-tubule cell, which gradually replaces the DTC-cell, is regarded as the beginning of the connecting tubule. The major portion of the connecting tubule is lined by connecting-tubule cells and intercalated cells. The first appearance of the principal cell between them defines the beginning of the cortical collecting duct, which in the medullary ray is lined by principal and intercalated cells only. Each cell type is described in detail and discussed in relation to the assumed function of the tubular segments. Interspecies differences in the cellular composition of the transitional zone from the nephron to the collecting duct are discussed in relation to the different organization of the collecting duct system.