Aquaporin-1 Is not Expressed in Descending Thin Limbs of Short-Loop Nephrons
Xiao-Yue Zhai*,,
Robert A. Fenton,
Arne Andreasen,
Jesper Skovhus Thomsen|| and
Erik I. Christensen*
Departments of * Cell Biology; Neurobiology; and || Connective Tissue Biology; The Water and Salt Research Center, Institute of Anatomy, University of Aarhus, Aarhus C, Denmark; and Department of Histology and Embryology, China Medical University, Shen Yang, China
Correspondence: Dr. Erik Ilsø Christensen, Department of Cell Biology, Institute of Anatomy, University of Aarhus, DK-8000 Århus C, Denmark. Phone: +45-89-42-30-57; Fax: +45-86-19-86-64; E-mail: eic{at}ana.au.dk
Received for publication January 16, 2007.
Accepted for publication June 19, 2007.
In mammalian kidneys, aquaporin-1 is responsible for water reabsorptionalong the proximal tubule and is also thought to be involvedin the concentration of urine that occurs in the medulla. Ithas been suggested, however, that aquaporin-1 is not expressedin the last part of the descending thin limbs of short loopnephrons in rats and mice, and its expression in this regionin humans has not been studied. We examined the expression ofaquaporin-1 and the urea transporter UT-A2 in serial sectionsof mouse nephrons in the inner stripe of the outer medulla usingimmunohistochemistry. In contrast to previous observations,we demonstrate a complete absence of aquaporin-1 along the entirelength of descending thin limbs of 90% of short loop nephrons.Conversely, as expected, we identified aquaporin-1 in proximaltubules, descending thin limbs of long loop nephrons, and medullarydescending vasa recta. We also observed this abrupt transitionfrom aquaporin-1–positive proximal tubules to aquaporin-1–negativedescending thin limbs of short loop nephrons in sections ofhuman and rat kidneys. UT-A2 was restricted to the last 28%to 44% of the descending thin limbs of all short loop nephrons.Because the majority of nephrons are of the short loop variety,our findings suggest that the mechanisms of water transportin the descending thin limbs of short loop nephrons should bereevaluated. Likewise, the roles of aquaporin-1 and UT-A2 inthe countercurrent multiplier and water conversation may needto be readdressed.
Aquaporin-1 (AQP-1), first discovered by Agre and colleagues,1,2is the highly expressed water channel of the kidney proximaltubules,3 where it is thought to be responsible for constitutivewater reabsorption in this nephron segment. AQP-1 is also expressedin the descending thin limb (DTL) of Henle's loop, and, althoughthis localization is consistently generalized in numerous reviews(see, e.g., Nejsum4 and Verkman5) with a few exceptions,6 ithas been suggested that the last part of the DTL of short-loopnephrons (SLN) in rats and mice does not express AQP-17,8; however,such a distinction was not made in a study on human kidney.9The exact localization of AQP-1 in the proximal tubule and DTLis thought to be essential for the urinary concentrating mechanismand, in the case of the latter, specifically the process ofcountercurrent multiplication. This is underlined by the observationthat the water permeability in proximal tubules from AQP-1 knockoutmice was reduced five-fold compared with normal mice.10 Similarly,an 8.5-fold reduction in water transport was observed in DTLof AQP-1 knockout mice.11 Together, these reduced water permeabilitiescontribute to the major urinary concentrating defect observedin these mice.
Histotopographic electron microscopic analyses12,13 and computer-assistedtracing of renal nephrons14 have determined that the DTL ofSLN and long-loop nephrons (LLN) consist of four types of epithelialcells. Type 1 epithelium is characterized by a flat, simpleepithelium that supposedly covers the entire thin limbs of SLN,whereas the other three epithelial types are distributed successivelyin the DTL and ascending thin limbs of LLN. Studies of microdissectedrat renal tubules15 estimated that the number of AQP-1 moleculesper tubule length in short-loop DTL type 1 epithelium was similarto that of segment 1 and segment 2 proximal tubules but seven-foldlower than that of type 2 epithelium of DTL of LLN.
The urea transporter UT-A2 is expressed in rats and mice inthe inner stripe of the outer medulla (ISOM), where it is localizedto the lower portions of the DTL of SLN8,16 and, under prolongedantidiuretic conditions, in the base of the inner medulla, whereit is localized to the DTL of LLN.8 UT-A2 is proposed to beinvolved in urea recycling, a process whereby urea that is reabsorbedfrom the inner medullary collecting duct (IMCD) is secretedinto the descending loop of Henle, causing it to be returnedto the collecting duct lumen with the flow of tubule fluid,thereby maintaining a high medullary interstitium urea concentrationand allowing the formation of a concentrated urine.17 In thisstudy, we extensively examined the nephron location of bothAQP-1 and UT-A2, based to a large extent on traced renal tubulesfrom serial sections of mouse kidneys. Our results show thatthere is a complete absence of AQP-1 labeling in the type 1epithelium of short-loop DTL of mouse kidneys, a finding thatis also revealed in rat and human kidneys. These findings mayhave major implications on the understanding of the urinaryconcentrating mechanism, specifically the role of AQP-1 in countercurrentmultiplication mechanisms in the ISOM.
AQP-1
No labeling for AQP-1 was detected in the thin limbs of thetraced mouse SLN (n = 147 in three kidneys) with type 1 epithelium.In contrast, abundant labeling was observed in the vascularbundles in descending vasa recta and in the inter bundle regionsin thin limbs of the traced LLN (n = 50 in three kidneys; Figure 1).This observation was further underpinned by a series of lightand electron micrographs demonstrating the transitions fromAQP-1–labeled proximal tubules into either nonlabeled,type 1 epithelium or into labeled type 2 epithelium of the LLN(Figures 2 and 3. Exceptions for these observations are demonstratedin Figure 4, where a labeled proximal tubule continues intoa labeled relatively thick, type 2–like epithelium that,after a distance (90 to 400 µm), is replaced with a nonlabeledthin type 1 epithelium. These nephrons were identified by tracingas the longest SLN with their bends located close to the borderbetween the outer and inner medulla, and they constitute approximately10% of all traced SLN.14
Figure 1. Three selected levels from the outer (A), middle (B), and inner part (C) of the ISOM of a mouse kidney showing Epon sections labeled for AQP-1. The DTL of LLN and SLN were identified by tracing and are indicated with either an L or an s. The areas composed of the thin-walled structures constitute the so-called vascular bundles, whereas the areas with the thick-walled tubules placed in between are interbundle regions. The labeled, thin-walled structures in the vascular bundles are the descending vasa recta. None of the traced SLN in the vascular bundles shows labeling for AQP-1, whereas the traced DTL of LLN in the interbundle region are labeled. In the inner part of the ISOM (C), the vascular bundles are smaller than at the earlier levels (A and B). Magnification, x80.
Figure 2. Electron micrographs of Epon sections demonstrating ultrastructurally two tubular epithelial transitions of mouse nephrons. (A) Transition from a proximal tubule to type 1 epithelium of the DTL of an SLN. The left inset shows the same transition at light microscope level from the labeled proximal tubule to the nonlabeled DTL of the SLN. The right inset shows ultrastructurally the transition at a higher magnification. (B) Transition from a proximal tubule to type 2 epithelium of the DTL of an LLN. The inset shows the transition at light microscope level of the labeled proximal tubule continuing into the DTL of the LLN. Bars = 5 µm in A and 2 µm in B. Magnifications: x600 in left inset in A; x5600 in right inset in A; x1000 in inset in B.
Figure 3. AQP-1 immunocytochemistry of ultrathin cryosections from a mouse kidney. (A) Immunocytochemistry for AQP-1 of two tubular transitions. (B) One of the transitions from A (top box). Gold labeling for AQP-1 in the proximal tubule at the brush border (BB) and at the basolateral membrane stops at the transition (arrow) into the nonlabeled type 1 epithelium. (C) The other transition from A (bottom box). The gold labeling for AQP-1 in the proximal tubule BB continues (arrow) into the type 2 epithelium at both the apical and basolateral membranes. Bars = 5 µm in A and 1 µm in B and C.
Figure 4. Electron micrograph showing the type 2–like epithelium that forms the initial part of the DTL of one of the longest traced mouse SLN (type 3 bend). The section was obtained 55 µm from the transition of the proximal tubule into the DTL. The insets show AQP-1 labeling of the same tubule at the transition (top left), approximately 110 µm from the transition (top middle), and 560 µm from the transition (top right). Bar - 2 µm. Magnifications: x100 in top left; x200 in top middle and right.
The observation that the type 1 epithelium of DTL of SLN doesnot express AQP-1 was further substantiated by investigationsof serial paraffin sections from the ISOM of rat and human kidneys.In many images, AQP-1 labeling stopped abruptly at the transitionfrom proximal straight tubules to DTL covered with a very thinepithelium (probably type 1 cells; Figure 5). In contrast, transitionsfrom proximal tubules to DTL with a taller epithelium (probablytype 2 cells) showed continuous labeling for AQP-1 similar towhat was observed in mice. In addition, double-labeling experimentsfor AQP-1 and CD34 (expressed in vascular endothelial cells)revealed many unlabeled, thin-walled structures in the interbundleregions of human kidneys (Figure 6), which most likely representDTL of SLN. In rats, there are more continuous labeled transitionsfrom proximal tubules to DTL covered with thick epithelium (probablytype 2) than in mice. Whether these labeled type 2–likeepithelia constitute either DTL of LLN or beginnings of DTLof the longest SLN cannot be established without performinga tracing of the nephrons. Taken together, these findings stronglysuggest that, similar to the mouse, the DTL of SLN in rat andhuman kidney are also mainly devoid of AQP-1.
Figure 5. Sections labeled for AQP-1 from rat (A; Epon) and human kidney (B; paraffin). Many cross and elongated tubular profiles exhibit two kinds of transitions: Some from labeled proximal tubules into nonlabeled, thin-walled DTL (probably type 1 epithelium, arrows), whereas others are continuously labeled from the proximal tubules into the relatively thick-walled DTL (probably type 2 epithelium, arrowheads). Magnifications: x150 in A; x300 in B.
Figure 6. Paraffin section double labeled for AQP-1 (brown) and CD4 (pink, vascular endothelial marker) in a human kidney. L, labeled, relatively thick DTL of the LLN; S, nonlabeled DTL of the SLN; avr, ascending vasa recta. The arrows mark the double-labeled descending vasa recta. Magnification, x150.
To investigate whether hydration status had an effect on AQP-1localization, we performed immunolocalization studies on kidneysfrom mice that were either water loaded or water restricted.Under our experimental conditions, the urine osmolality of thewater-loaded group was 412 ± 82 mOsmol/kg H2O (n = 3)compared with 3213 ± 276 mOsmol/kg H2O (n = 3) in water-restrictedmice. Under these conditions, there was still no AQP-1 labelingof DTL of SLN, whereas labeling of other parts of the nephronsand the descending vasa recta was unchanged (data not shown).
UT-A2
The terminal segment of the DTL of all mouse SLN expresses UT-A2.These labeled segments are located in the inner part of theISOM (Figure 7), confirming previous studies.8 Uninterruptedlabeling continued to the transition between the thin limb andthe thick ascending limb (Figure 8). Zhai et al.14 classifiedmouse SLN into three different types according to their bends.For each of the three types of SLN, the length of the UT-A2–labeledsegment was determined in one kidney to be 205 ± 55 µm(n = 13) for nephrons with type 1 bends, 238 ± 91 µm(n = 9) for nephrons with type 2 bends, and 353 ± 119µm (n = 19) for nephrons with type 3 bends. SLN with eithertype 1 or type 2 bends originate from glomeruli located an averageof 200 and 300 µm from the renal surface, respectively,and with bends located at the middle level of the ISOM. SLNwith type 3 bends originate from nephrons with glomeruli locatedan average of 500 µm from the renal surface and with bendslocated at various levels of the inner half of the ISOM. Type1 bends are covered with DTL epithelium continuing a short distanceinto the ascending limb, type 2 bends have an epithelial transitionfrom DTL into thick ascending limb at the bend, and type 3 bendshave a prebend thick ascending limb segment of varying length.14For all three types of nephrons, UT-A2 labeling begins approximately470 µm (232 to 623 µm) from the proximal straighttubule. The fraction of UT-A2–labeled thin-limb epitheliumconstituted 28, 35, and 44% of the total DTL length for thethree types of SLN.
Figure 7. Two selected levels from the middle (A) and inner (B) part of the ISOM of a mouse kidney showing Epon sections labeled for UT-A2. The two levels correspond to the two levels (consecutive sections) shown in Figure 1, B and C. The traced DTL of SLN marked with S in the lumen all are labeled for UT-A2 in the vascular bundles, whereas the traced DTL of the LLN marked with L in the lumen in the interbundle regions are not labeled for UT-A2. Magnification, x80.
Figure 8. Micrograph of an Epon section showing the transition (arrows) from UT-A2–labeled DTL into the nonlabeled, thick ascending epithelium of an SLN of the mouse kidney. Magnification, x200.
Labeling of serial paraffin sections from the inner part ofthe ISOM of rat and human kidneys revealed individually exclusivelabeling for AQP-1 in DTL of LLN and descending vasa recta andfor UT-A2 in DTL of SLN (data not shown) as demonstrated previouslyin rat and mouse kidneys.8 Thus, in rat and human kidney, short-loopDTL type 1 epithelia do not express AQP-1 but only UT-A2 inthe terminal segment of the DTL, which is in agreement withour findings in mice.
The major finding of this study is that DTL of SLN in mouse,rat, and human kidney to a major extent are devoid of AQP-1under physiologic, hydrated, or dehydrated conditions. Thisdiscovery is potentially of great importance for our understandingof the urinary concentrating mechanism, particularly becauseprevious studies have considered that DTL of both SLN and LLNare highly permeable to water as a result of AQP-1 expression.
Our extensive analysis of AQP-1 expression from 147 DTL of SLNin three different mouse kidneys through their entire lengthdemonstrates conclusively that in this species, DTL type 1 epitheliumof SLN does not express AQP-1. This observation was furthersubstantiated by demonstrating direct transitions of AQP-1–labeledpars recta of proximal tubules into unlabeled DTL. The nephronsdemonstrating labeling of the first approximately 90 to 400µm of the DTL that then transition into unlabeled DTLcells are very long SLN that have their bends close to the borderbetween the outer and inner medulla. These very long SLN constitutea minor population of approximately 10% of all SLN.14 Thesenephrons have a short initial segment covered with type 2–likeepithelial DTL cells very similar to the type 2 epithelium coveringthe DTL of LLN. Our findings in rat and human kidneys demonstratethat the lack of AQP-1 in DTL of SLN is not limited to micebut is probably a general feature of other species; however,in rats, the transitions from pars recta proximal tubule intotype 1 epithelium seem to be less frequent than in mice.
Although the majority of our immunolabeling was performed onEpon sections, the AQP-1 labeling observed in these sectionswas not different from labeling of paraffin sections. Furthermore,the otherwise strong labeling of other nephron segments andof the descending vasa recta precludes the possibility thatthe lack of labeling of the DTL of SLN could be due to the preparationand treatment of the tissue.
Our findings are in contrast to previous studies that reportedAQP-1 expression in SLN.7–9 Also, in a study by Maedaet al.15 on microdissected rat renal tubules quantifying AQP-1abundance by ELISA, the number of AQP-1 molecules in DTL ofSLN was estimated to be equally high per tubule length as segment1 and 2 proximal tubules but a factor 7 less than that of type2 epithelium of DTL of LLN. The most plausible reasons for thesediscrepancies may be difficulties in distinguishing short-loopDTL from descending vasa recta and long-loop DTL morphologicallyand during microdissection.
Anatomically, the ratio of SLN to LLN is 82:18 in mice,14 70:30in rats, and 85:15 in humans (for a review, see Bankir and deRouffignac18). Our data suggest that approximately 82% of allDTL in mice do not possess AQP-1. This figure may be somewhatlower for rats in part because of the difference in the ratiobetween SLN and LLN. Also, it is not known whether the rat hasthe very long type 3 SLN similar to those found in mice.14 Whereasthe AQP-1–deficient DTL are located inside the vascularbundles in mice and at the periphery in rats, the DTL of LLNare located in the inter bundle regions of the ISOM in bothspecies.14,19,20 In human kidney, both types of DTL are locatedin the inter bundle regions.21
From a physiologic perspective, the lack of AQP-1 in short DTLis intriguing. As originally described by Kuhn and Ramel,22the urinary concentrating mechanism depends on the process ofcountercurrent multiplication, whereby the loops of Henle serveas a countercurrent multiplier system that generates and maintainsa high medullary osmotic gradient and allows the osmotic extractionof water from the collecting duct system (see Knepper and Gamba23for review). For the countercurrent multiplication to work effectively,osmotic equilibration in the descending limbs must occur andthe tubule fluid of the DTL must become progressively more concentratedas it flows toward the papilla, a feature confirmed by earlymicropuncture studies.24 The mechanism by which osmotic equilibrationoccurs could be by solute entry, water exit, or both (discussedby Levine et al.25). Indeed, numerous studies in several speciesusing the isolated perfused tubule technique have shown thatthe water permeability of DTL of LLN is extremely high as aresult of the high abundance of AQP-111,26,27; however, to ourknowledge, only one study measuring the water permeability ofDTL of SLN has been performed,28 presumably because of the thin/fragilewall of the type 1 epithelium making intact tubules virtuallyimpossible to isolate manually. This single study, performedin hamsters, indicated that the DTL of SLN are indeed permeableto water, but their osmotic water permeability is much lowerthan that of the DTL of LLN. Importantly, in these experiments,the authors stipulated that they could not discount the possibilitythat one or more of their perfused "type 1 epithelia" are indeed"type 2 epithelia"; therefore, the actual osmotic water permeabilityof the DTL of SLN may be even lower than actually reported.These results lead us to speculate that the lower water permeabilityobserved in DTL of SLN is due to a lack of AQP-1. In addition,our results suggest, in part, that both the mechanism of soluteentry and water removal may occur in the DTL of SLN and thatosmotic equilibration can occur as a result of solute entrycoupled to a small amount of water efflux independent of AQP;however, further studies examining the water permeability ofSLN are required to clarify this issue, although such investigationsare likely to be extremely technically challenging.
If AQP-1 is not expressed in DTL of SLN but the loops are permeableto water,28 then what accounts for this observed water permeability?Studies by Chou et al.11 clearly showed that even in the absenceof AQP-1, DTL of LLN have a relatively high water permeability,and this may also be the case for DTL of SLN. One possible explanationfor the observed water permeability28 is that another AQP waterchannel localizes to this nephron segment. Six other AQP havebeen localized to the mammalian nephron, but it is unlikelythat any of these are suitable candidates because they all havebeen extensively studied; however, the possibility exists thatanother, as-yet-unidentified AQP localizes to the DTL of SLN,and further studies are required to investigate this. Anotherpossibility for such a high water permeability is a paracellularpathway, but because the tight junctions of type 1 epitheliumare characterized by a deep junctional complex12 containingseveral ramified junctional strands, characteristic of "tight" tight junctions, this pathway seems unlikely to be responsible.The final potential explanation is water movement through thelipid bilayer; however, because type 1 epithelia do not containabundant convoluted apical and basolateral membranes, like thetype 2 epithelium, the contribution of this pathway to membranepermeability remains speculative. In contrast, if the DTL ofSLN are not permeable to water (because only one study has everreported such findings), then further studies are required toinvestigate the nature of solute concentration along the descendinglimb.
Studies of mathematical models have given great insight intothe role of a short portion of the DTL of SLN that is waterimpermeable but urea permeable. One such recent study, by Laytonet al.,29 clearly showed that a water-impermeable segment ofthe DTL of SLN is actually beneficial to maximal urinary concentratingability. In their studies, the replacement of a short water-impermeableand highly urea-permeable segment (so-called SDL2 segment) witha highly water-permeable and moderately urea-permeable segment(so-called SDL1 segment) reduced collecting duct osmolality,outer medullary free water absorption, and efficiency. The authors’explanation for their findings was that the presence of thewater-impermeable segment increases collecting duct fluid osmolalitynear the outer medulla–inner medulla boundary by eliminatingthe load that would otherwise arise from fluid absorbed fromthe terminal portion of the SDL. With respect to our findings,it would be interesting to model mathematically the same processesin a DTL of an SLN that has water permeability equivalent tothat observed in DTL of LLN of AQP-1 knockout mice to examinethe overall effect on concentrating ability.
In addition to our findings with respect to AQP-1, our studiesdetermined that the facilitative urea transporter UT-A2 is expressedin all three types of SLN. On the basis of our tracing studies,the expression begins approximately 470 µm from the transitionof the proximal straight tubule into the DTL and continues tothe transition between the thin limb and the thick ascendinglimb. For water extraction from the vasopressin-sensitive collectingduct, a corticomedullary osmolality gradient must exist withinthe kidney, with a maximum osmolality at the tip of the innermedulla. In the inner medulla, this gradient is composed mainlyof urea,30 which is passively reabsorbed from the IMCD via theurea transporters UT-A1 and UT-A3. Although the process of countercurrentexchange minimizes dissipation of urea from the inner medullaryinterstitium, it cannot completely eliminate urea loss becausethe volume flow rate of blood in the ascending vasa recta exceedsthat in the descending vasa recta. Thus, in addition to countercurrentexchange, urea recycling is believed to provide an importantmeans of maintaining a high level of urea in the renal innermedulla.17 Urea recycling occurs when the urea that is reabsorbedfrom the IMCD is re-secreted into the loop of Henle, causingit to be returned to the collecting duct lumen with the flowof tubule fluid. A major element of urea secretion into Henle'sloop is believed to occur via transfer from the vasa recta (containinganother facilitative urea transporter, UT-B) to the DTL of SLNin the vascular bundles of the outer medulla, where these twostructures are closely associated.14,20 Our localization ofUT-A2 suggests that intrarenal urea recycling must occur ina very specific location, although recent findings from knockoutmice studies suggest that urea secretion into the DTL of Henle'sloop does not seem to be as important in medullary urea accumulationas previously believed.31
This study has demonstrated that the majority of DTL of SLNin mouse, rat, and human kidneys do not express AQP-1, contraryto the general opinion. These findings are of great importancefor the understanding of the urine concentrating mechanisms.We have further obtained data giving the exact lengths of theexpression of UT-A2 in the same part of the nephron, data ofimportance for the mathematical modeling of transport mechanismsin the kidney.
Preparation of Renal Tissue
The tissue preparation of the mouse kidneys has previously beendescribed in detail.14 Briefly, kidneys from male C57/BL/6Jmice, 25 g body weight, were fixed by perfusion through theabdominal aorta with 1% glutaraldehyde in 0.06 M sodium cacodylatebuffer and 4% hydroxyethyl starch. Tissue blocks cut perpendicularto the longitudinal axis of the kidney were postfixed for 1h in 1% OsO4 and embedded in Epon 812. From each kidney, a totalof 2500, 2.5-µm-thick consecutive sections were obtainedfrom the surface to the papillary tip and stained with toluidineblue. Renal tissue was also obtained from rats or mice fixedby perfusion as described or with 2% paraformaldehyde and embeddedin either Epon or paraffin or frozen in liquid nitrogen forcryosectioning, respectively, by standard methods. In addition,renal sections were obtained from paraffin-embedded normal humankidneys (autopsy materials). All animal experiments were carriedout in accordance with provisions for the animal care licenseprovided by the Danish National Animal Experiments Inspectorate.Image recordings, alignment, and digital tracing were performedas described previously in detail.14
Effect of Hydration Status on AQP-1 Expression
Six mice were housed individually in normal cages and receivedfree access to drinking water and pelleted diet for severaldays before experimental manipulation. Subsequently, the micereceived a fixed daily ration of gelled diet containing either2 ml of deionized water (water restricted) or 8 ml of deionizedwater (water loaded) as described previously.32 The mice didnot have access to supplemental drinking water during this period.After 48 h, spot urine was collected, urine osmolalities weredetermined using a vapor pressure osmometer, and kidneys wereprocessed for Epon embedding as detailed previously.
Antibodies
A commercial rabbit anti–AQP-1 (AB3065; Chemicon Int.,Temecula, CA) was used at a dilution of 1:400. Rabbit anti–UT-A28was used at dilutions of 1:200 to 1:800. Mouse anti–AQP-1(ab6994) was obtained from Abcam (Cambridge, UK), and mouseanti-human CD34 was purchased from Immunotech (Marseilles, France)and used at dilutions of 1:100 and 1:400, respectively.
Immunohistochemistry
A total of 20, 2.5-µm-thick Epon sections (50 µmbetween the sections) from the start of the ISOM until the innermedulla were re-embedded in Epon and sectioned into two 0.5-µm-thicksections for each of the 2.5-µm-thick sections. Thus,the sections span the entire length of the DTL of SLN. The re-embeddedsections originated from one of the three kidneys that havebeen three-dimensionally reconstructed and traced by Zhai etal.14 The sections were pretreated as described previously,33and one glass was incubated with anti–AQP-1 and the otherwith anti–UT-A2, followed by incubation with peroxidase-conjugatedsecondary anti-rabbit antibody (DAKO A/S, Glostrup, Denmark).Sections from similar selected zones from the two other micetraced by Zhai et al. were treated identically. Consecutiveparaffin sections from rat and human kidneys were incubatedsimilarly with the same antibodies. Epon and paraffin sectionsfrom rat and paraffin sections from human kidneys were examinedat the border of the inner and outer stripes of the outer medullato visualize the transitions from the proximal tubule to theDTL. The sections were examined using a Leica (Herlev, Denmark)DMR microscope equipped with a Leica DFC320 camera. Images wereacquired by the Leica TFC TWAIN 6.1.0 driver and processed usingAdobe Photoshop 8.0 (Adobe, San José, CA). The areasexamined from the traced kidneys were selected so as to includeas many of the traced tubules as possible.
Length Measurements
The lengths of the DTL of mouse SLN labeled for UT-A2 were calculatedas the sum of the Euclidian distance starting at the initiallabeling until the transition into the thick ascending limbas described previously in detail.14
Electron Microscopy
On the basis of the tracing and immunolabeling, sections representingthe transitions between different cell types were selected forultrastructural analysis. These sections were re-embedded inEpon, sectioned in 50-nm-thick sections, and observed in a PhilipsCM 100 electron microscope. Ultrathin cryosections, 80 nm thick,incubated with rabbit anti–AQP-1 and subsequently withgoat anti-rabbit 10-nm gold particles (British BioCell Int.,Cardiff, UK), were likewise examined in the electron microscope.
The work was supported in part by the Danish Medical ResearchCouncil, the University of Aarhus, the Novo-Nordisk Foundation,the Biomembrane Research Center, National Natural Science Foundationof China (contract 30640011), and the European Commission (EUFramework Program 6, EureGene, contract 05085). R.A.F. is supportedby a Marie Curie Intra-European Fellowship, the Carlsberg Foundation(Carlsbergfonden), and the Danish National Research Foundation(Danmarks Grundforskningsfond).
The skilful technical assistance of Pia Nielsen, Hanne Sidelmann,and Inger Kristoffersen is gratefully acknowledged. We are thankfulto Dr. Mark A. Knepper (National Heart, Lung, and Blood Institute,Bethesda, MD) for helpful discussions.
Footnotes
Published online ahead of print. Publication date availableat www.jasn.org.
See related editorial, "Aquaporin 1, Urea Transporters, andRenal Vascular Bundles," on pages 2798–2800.
Preston GM, Agre P: Isolation of the cDNA for erythrocyte integral membrane protein of 28 kilodaltons: Member of an ancient channel family.
Proc Natl Acad Sci U S A 88
: 11110
–11114, 1991[Abstract/Free Full Text]
Preston GM, Carroll TP, Guggino WB, Agre P: Appearance of water channels in Xenopus oocytes expressing red cell CHIP28 protein.
Science 256
: 385
–387, 1992[Abstract/Free Full Text]
Nielsen S, Smith BL, Christensen EI, Knepper MA, Agre P: CHIP28 water channels are localized in constitutively water-permeable segments of the nephron.
J Cell Biol 120
: 371
–383, 1993[Abstract/Free Full Text]
Nejsum LN: The renal plumbing system: Aquaporin water channels.
Cell Mol Life Sci 62
: 1692
–1706, 2005[CrossRef][Medline]
Verkman AS: Roles of aquaporins in kidney revealed by transgenic mice.
Semin Nephrol 26
: 200
–208, 2006[CrossRef][Medline]
Nielsen S, Frokiaer J, Marples D, Kwon TH, Agre P, Knepper MA: Aquaporins in the kidney: From molecules to medicine.
Physiol Rev 82
: 205
–244, 2002[Abstract/Free Full Text]
Nielsen S, Pallone T, Smith BL, Christensen EI, Agre P, Maunsbach AB: Aquaporin-1 water channels in short and long loop descending thin limbs and in descending vasa recta in rat kidney.
Am J Physiol 268
: F1023
–F1037, 1995[Medline]
Wade JB, Lee AJ, Liu J, Ecelbarger CA, Mitchell C, Bradford AD, Terris J, Kim GH, Knepper MA: UT-A2: A 55-kDa urea transporter in thin descending limb whose abundance is regulated by vasopressin.
Am J Physiol Renal Physiol 278
: F52
–F62, 2000[Abstract/Free Full Text]
Maunsbach AB, Marples D, Chin E, Ning G, Bondy C, Agre P, Nielsen S: Aquaporin-1 water channel expression in human kidney.
J Am Soc Nephrol 8
: 1
–14, 1997[Medline]
Schnermann J, Chou CL, Ma T, Traynor T, Knepper MA, Verkman AS: Defective proximal tubular fluid reabsorption in transgenic aquaporin-1 null mice.
Proc Natl Acad Sci U S A 95
: 9660
–9664, 1998[Abstract/Free Full Text]
Chou CL, Knepper MA, Hoek AN, Brown D, Yang B, Ma T, Verkman AS: Reduced water permeability and altered ultrastructure in thin descending limb of Henle in aquaporin-1 null mice.
J Clin Invest 103
: 491
–496, 1999[Medline]
Bachmann S, Kriz W: Histotopography and ultrastructure of the thin limps of the loop of Henle in the hamster.
Cell Tissue Res 225
: 111
–127, 1982[CrossRef][Medline]
Takahashi-Iwanaga H, Iwata Y, Adachi K, Fujita T: The histotopography and ultrastructure of the thin limb of the Henle's loop: A scanning electron microscopic study of the rat kidney.
Arch Histol Cytol 52
: 395
–405, 1989[Medline]
Zhai XY, Thomsen JS, Birn H, Kristoffersen IB, Andreasen A, Christensen EI: Three-dimensional reconstruction of the mouse nephron.
J Am Soc Nephrol 17
: 77
–88, 2006[Abstract/Free Full Text]
Maeda Y, Smith BL, Agre P, Knepper MA: Quantification of aquaporin-CHIP water channel protein in microdissected renal tubules by fluorescence-based ELISA.
J Clin Invest 95
: 422
–428, 1995[Medline]
Fenton RA, Stewart GS, Carpenter B, Howorth A, Potter EA, Cooper GJ, Smith CP: Characterization of mouse urea transporters UT-A1 and UT-A2.
Am J Physiol Renal Physiol 283
: F817
–F825, 2002[Abstract/Free Full Text]
Knepper MA, Roch-Ramel F: Pathways of urea transport in the mammalian kidney.
Kidney Int 31
: 629
–633, 1987[Medline]
Bankir L, de Rouffignac C: Urinary concentrating ability: Insights from comparative anatomy.
Am J Physiol 249
: R643
–R666, 1985[Medline]
Kriz W, Schnermann J, Koepsell H: The position of short and long loops of Henle in the rat kidney.
Z Anat Entwicklungsgesch 138
: 301
–319, 1972[CrossRef][Medline]
Kriz W, Koepsell H: The structural organization of the mouse kidney.
Z Anat Entwicklungsgesch 144
: 137
–163, 1974[CrossRef][Medline]
Kriz W: Structural organization of the renal medulla: Comparative and functional aspects.
Am J Physiol 241
: R3
–R16, 1981[Medline]
Kuhn W, Ramel A: Active transport of salt as a possible (and probable) individual effect in the urine concentration in the kidney.
Helv Chim Acta 42
: 628
–660, 1959[CrossRef]
Knepper MA, Gamba G: Urine concentration and dilution. In:
The Kidney, 7th Ed., edited by Brenner BM, Philadelphia, Saunders, 2004
, pp 599
–636
Gottschalk CW, Mylle M: Micropuncture study of the mammalian urinary concentrating mechanism: Evidence for the countercurrent hypothesis.
Am J Physiol 196
: 927
–936, 1959[Abstract/Free Full Text]
Levine DZ, Roinel N, de Rouffignac C: Flow-correlated influx of K, Ca P, and Mg during continuous microperfusion of the loop of Henle in the rat.
Kidney Int 22
: 634
–639, 1982[CrossRef][Medline]
Imai M, Yoshitomi K: Heterogeneity of the descending thin limb of Henle's loop.
Kidney Int 38
: 687
–694, 1990[Medline]
Miwa T, Imai M: Flow-dependent water permeability of the rabbit descending limb of Henle's loop.
Am J Physiol 245
: F743
–F754, 1983[Medline]
Imai M, Hayashi M, Araki M: Functional heterogeneity of the descending limbs of Henle's loop. I. Internephron heterogeneity in the hamster kidney.
Pflugers Arch 402
: 385
–392, 1984[CrossRef][Medline]
Layton AT, Layton HE: A region-based mathematical model of the urine concentrating mechanism in the rat outer medulla. II. Parameter sensitivity and tubular inhomogeneity.
Am J Physiol Renal Physiol 289
: F1367
–F1381, 2005[Abstract/Free Full Text]
Jarusch KH, Ullrich KJ: Studies on the problem of urine concentration and dilution; distribution of electrolytes (sodium, potassium, calcium, magnesium, anorganic phosphate), urea amino acids and exogenous creatinine in the cortex and medulla of dog kidney in various diuretic conditions.
Pflugers Arch 262
: 537
–550, 1956[CrossRef][Medline]
Uchida S, Sohara E, Rai T, Ikawa M, Okabe M, Sasaki S: Impaired urea accumulation in the inner medulla of mice lacking the urea transporter UT-A2.
Mol Cell Biol 25
: 7357
–7363, 2005[Abstract/Free Full Text]
Fenton RA, Chou CL, Stewart GS, Smith CP, Knepper MA: Urinary concentrating defect in mice with selective deletion of phloretin-sensitive urea transporters in the renal collecting duct.
Proc Natl Acad Sci U S A 101
: 7469
–7474, 2004[Abstract/Free Full Text]
Zhai XY, Kristoffersen IB, Christensen EI: Immunocytochemistry of renal membrane proteins on epoxy sections.
Kidney Int 72
: 731
–735, 2007[CrossRef][Medline]
Related Article
Aquaporin 1, Urea Transporters, and Renal Vascular Bundles
Thomas L. Pallone
J. Am. Soc. Nephrol. 2007 18: 2798-2800.
[Full Text][PDF]
This article has been cited by other articles:
U. Hasler, V. Leroy, P.-Y. Martin, and E. Feraille Aquaporin-2 abundance in the renal collecting duct: new insights from cultured cell models
Am J Physiol Renal Physiol,
July 1, 2009;
297(1):
F10 - F18.
[Abstract][Full Text][PDF]
D. Bockenhauer, M. Cruwys, R. Kleta, L.F. Halperin, P. Wildgoose, T. Souma, N. Nukiwa, S. Cheema-Dhadli, C.K. Chong, K.S. Kamel, et al. Antenatal Bartter's syndrome: why is this not a lethal condition?
QJM,
December 1, 2008;
101(12):
927 - 942.
[Abstract][Full Text][PDF]
Top
Abstract
Introduction
