Axial heterogeneity in basolateral AQP2 localization in ratkidney: effect of va - 《American Journal of Physiology》美国生理学杂志干细胞之家

干细胞之家 - 中国干细胞行业门户第一站 › 干细胞之家论坛 › 干细胞文献资源库 › 《American Journal of Physiology》美国生理学杂志 › Axial heterogeneity in basolateral AQP2 localization ...



免疫细胞治疗专区	欢迎关注干细胞微信公众号

12 3 4 5 6 7 8 9 10 ... 29 下一页

返回列表

查看: 692197\|回复: 285	go Axial heterogeneity in basolateral AQP2 localization in ratkidney: effect of va [复制链接]

轻羽

新手上路

Rank: 1

积分: 0
威望: 0
包包: 0

楼主

发表于 2009-4-21 13:35 |只看该作者 |倒序浏览 |打印

作者：Birgitte MønsterChristensen, WeidongWang, JørgenFrøkiær,  SørenNielsen,作者单位：1 The Water and Salt Research Center, Institute of Anatomy, and Instituteof Experimental Clinical Research, University of DK-8000 Aarhus,Denmark " F- r+ H5 D' T  d$ e7 Y
   ' m$ n; `, X) c# I  P

      % z7 h7 k, m# H+ B- i% H% a3 }" u

         ' j- z" `! L/ U3 f5 T1 Z! [5 I

         ! Z# e& b! J. B% @2 ]' P
         * X  J4 |( W$ _8 F

      . J$ i) K% Z; J. a0 C/ T1 U; O


      【摘要】" y# K/ O" q9 H
   The purpose of the present studywas to examine whether there is axial heterogeneity in the basolateralplasma membrane (BLM) localization of AQP2 and whether alteredvasopressin action or medullary tonicity affects the BLM localizationof AQP2. Immunocytochemistry and immunoelectron microscopy revealedAQP2 labeling of the BLM in connecting tubule (CNT) cells and innermedullary collecting duct (IMCD) principal cells in normal rats andvasopressin-deficient Brattleboro rats. In contrast there was littlebasolateral AQP2 labeling in cortical (CCD) and outer medullarycollecting duct principal cells. Short-termdesamino-Cys 1, D -Arg 8 vasopressin(dDAVP) treatment (2 h) of Brattleboro rats caused no increase in AQP2labeling of the BLM. In contrast, long-term dDAVP treatment (6 days) ofBrattleboro rats caused an increased BLM labeling in CNT, CCD, andIMCD. Treatment of normal rats with V 2 -receptor antagonistfor 60 min caused retrieval of AQP2 from the apical plasmamembrane. Moreover, AQP2 labeling of the BLM was unchanged in CNTand IMCD but increased in CCD. In conclusion, there is an axialheterogeneity in the subcellular localization of AQP2 with prominentAQP2 labeling of the BLM in CNT and IMCD. There was no increase in AQP2labeling of the BLM in response to short-term dDAVP. Moreover, acuteV 2 -receptor antagonist treatment did not cause retrieval ofAQP2 from the BLM. In contrast, long-term dDAVP treatment caused amajor increase in AQP2 expression in the BLM in CCD.
      【关键词】 aquaporin basolateral vasopressin/ R8 ?* {+ S# B" h# P, b
   INTRODUCTION6 d. I5 j5 ]5 E/ |. e" j9 I0 A

TWO-THIRDS OF THE WATER entering the collecting duct is reabsorbed in the corticalcollecting duct (CCD) due to osmotic equilibration with the nearisosmotic cortical interstitium. The remaining urine enters themedullary collecting duct where it equilibrates with thehyperosmotic interstitium, resulting in excretion of a concentrated urine. The vasopressin-induced water permeability as well as the translocation of aquaporin-2 (AQP2) from intracellular vesicles to theapical plasma membrane in collecting duct principal cells is wellcharacterized in the inner medullary collecting duct (IMCD) (17, 21, 26, 34, and for review, see Ref. 23 ), but virtually nothing is known about the trafficking of AQP2 in more proximal parts( 17, 23 ). AQP2 is also found in the basolateral plasma membrane in IMCD principal cells ( 22 ), but it is uncertainwhether vasopressin is involved in the basolateral targeting of AQP2. It is also unclear whether the same localization pattern of AQP2 isfound in the proximal part of the collecting duct system, i.e., whetherthere is an axial heterogeneity in the basolateral expression of AQP2along the connecting tubule (CNT) and collecting duct.& e. h' U) ?* H; ?- Q& Y7 h

Recent studies have attempted to address the mechanisms underlying thebasolateral targeting of AQP2. Vasopressin has been speculated to causethe basolateral targeting, since acute in vitro vasopressin treatmentof kidney slices from normal rats has been shown to be associated withbasolateral labeling of AQP2 ( 2 ). That oxytocin is also apotential candidate is shown by acute oxytocin treatment of normalrats, which induces distinct basolateral targeting of AQP2 in additionto apical targeting ( 10 ). In vitro studies performedby van Balkom et al. ( 33 ) have also indicated thathypertonicity may directly or indirectly induce targeting of AQP2 tothe basolateral plasma membrane domains in AQP2-transfectedMadin-Darby canine kidney (MDCK) cells. Finally, in some preparations,AQP2-transfected LLC-PK 1 cells and primary cultured IMCDcells exhibited forskolin- or vasopressin-induced basolateral targetingof AQP2 ( 11, 16 ), whereas in other preparations MDCK cells and cells from a human cortical collecting duct cell linetransfected with AQP2 showed only apical AQP2 targeting in response tovasopressin or forskolin stimulation ( 6, 31 ). Thus thereis a great deal of confusion as to the presence or absence of regulatedbasolateral AQP2 targeting and whether this has any relevance in vivo.
9 x$ x7 u1 Z7 k, c
Thus there is need for a rigorous study to define the subcellularlocalization (with a focus on the basolateral localization) of AQP2along the CNT and collecting duct segments and, moreover, to establishwhether there are major changes in the basolateral localization inconditions with altered vasopressin action. Specifically, in this studywe investigated the basolateral localization of AQP2 along the CNT andthe collecting duct in normal rats and in vasopressin-deficientBrattleboro rats using immunocytochemistry at the light- and electronmicroscopic level. We furthermore investigated whether conditions withaltered vasopressin action are associated with changes in AQP2 levelsin the basolateral plasma membrane in the CNT and in the differentsegments of the collecting duct. The following rat models were used: 1 ) vasopressin V 2 -receptor antagonist (1 h)-treated normal rats and 2 ) short-term (2 h) and long-term(6 days) desamino-Cys 1, D -Arg 8 vasopressin (dDAVP)-treated Brattleboro rats.

MATERIALS AND METHODS' F% P, Y5 e" q( j

Experimental Animals2 T. i3 S1 Q0 p  w5 }

Male Wistar rats were obtained from M & B (Ry, Denmark) andfemale homozygous Brattleboro rats from Harlan Netherlands (Horst, The Netherlands).
3 i* y/ g. O* _0 G' l" O
Experimental Protocols3 x1 g- I4 }0 u
$ q5 j- n: O  ~4 P3 k4 |- R" s% Y+ V
Protocol 1. (1-[4- N -tert-butylcarbamoyl)-2-methoybenzenesulfonyl]-5-ethoxy-3-spiro-[4-(2-morpholinoethoxy)cyclohexane] indol-2-onephosphate monohydrate; cis-isomer (SR-121463B) is a nonpeptide,V 2 -receptor antagonist (Sanofi-SynthélaboRecherche, Toulouse, France). SR-121463B was given via thefemoral vein to normal rats briefly anesthetized with halothane( n = 8). Control rats received saline( n = 8). Rats were placed in metabolic cages for 60 min, and anesthesia was then repeated. Kidneys were perfusion fixed forimmunocytochemistry, paraffin embedding (eight controls and eightSR-121463B-treated rats), and cryosubstitution (four controls and fourSR-121463B-treated rats).

Protocol 2. Brattleboro rats were treated subcutaneously with dDAVP (Sigma, Seelze,Germany; 500 ng in 500 µl saline/animal, n = 13). Control rats received subcutaneous saline ( n = 11). Rats were placed in metabolic cages, and urine output andwater intake were monitored. After 2 h of injection, anesthesiawas repeated, and kidneys were perfusion fixed for immunocytochemistry,paraffin embedding (eight controls and nine dDAVP-treated rats), andcryosubstitution (eight controls and nine dDAVP-treated rats).9 G0 h2 I6 A% I) ^$ P
" z3 z$ M" v9 i  j
Protocol 3. For long-term treatment of dDAVP, osmotic minipumps (Alzet, Palo Alto,CA) were filled with a solution of dDAVP and saline to give a dose of250 ng/day ( n = 6). Brattleboro rats were anesthetized, and the pump was placed subcutaneously. Controls were treated identically except that dDAVP was omitted from the solution( n = 5). Rats were put in metabolic cages with freeaccess to food and water. Urine output and water intake were monitoredfor 6 days. After 6 days, rats were anesthetized, and the kidneys were perfusion fixed for immunocytochemistry.3 D1 X6 c- z  m6 h0 P7 n5 n4 h# N
6 K) J. }$ [4 @  E3 b2 z1 s2 J
Immunocytochemistry

Fixation. Kidneys were perfusion fixed with 1 ) 2%paraformaldehyde in 0.01 M NaIO 4, 0.075 M L -lysin, 0.0375 M Na 2 HPO 4, pH 6.2; 2 ) 3% paraformaldehyde in 0.1 M sodium cacodylate, pH 7.4;or 3 ) 4% paraformaldehyde in 0.1 M sodium cacodylate, pH7.4, via the abdominal aorta. Kidneys within the same experiment wereidentically fixed. Kidneys were postfixed for 30 min in 0.1 M sodiumcacodylate. One kidney was subjected to paraffin embedding and onekidney for cryosubstitution.

Preparation of tissue for light and laser confocal microscopy. Before paraffin embedding, tissue blocks from whole kidneys weredehydrated in ethanol (2 h in 70, 96, and 99%, respectively) followedby incubation in xylene overnight. Paraffin sections (2 µm) were cuton a Leica RM 2126 microtome and dried overnight at 37°C. Sectionswere incubated with affinity-purified antibodies against rat or humanAQP2, and labeling was visualized by use of peroxidase-conjugatedsecondary antibody. Double labeling was also performed with polyclonalantibodies against rat AQP2 and monoclonal antibodies against mousecalbindin D-28K (Research Diagnostics, Flanders, NJ) to distinguishbetween CCD and CNT. In the rat, calbindin is abundantly expressed indistal convoluted tubule (DCT) and CNT, where it is involved in calciumreabsorption. Calbindin is reported to be present only in a fewcollecting duct cells in the superficial cortex in the rat ( 1, 28 ). The labeling was visualized with Alexa 488- and Alexa546-conjugated secondary antibodies (Alexa 488 anti-rabbit and Alexa546 anti-mouse, respectively). Light microscopy was carried out with aLeica DMRE microscope and Zeiss and Leica laser confocal microscopes.3 ]  ?; E  `* C! K  r

Preparation of tissue for immunoelectron microscopy. Tissue blocks prepared from the upper third of kidney inner medulla andcortex were infiltrated with 2.3 M sucrose for 30 min, mounted onholders, and rapidly frozen in liquid nitrogen. Frozen tissue blockswere subjected to cryosubstitution and Lowicryl HM20 embedding.Cryosubstitution was performed as previously described ( 20, 24 ). Ultrathin (80 nm) Lowicryl sections were cut on a Reichert Ultracut S and were preincubated with 0.05 M Tris, pH 7.4, 0.1% Triton-X-100 (TBST) containing 0.1% sodium borohydride and 0.05 M glycine followed by incubation with TBST containing 0.2% skimmedmilk. The preincubation was followed by incubation withaffinity-purified antibodies against rat or human AQP2 (sections frominner medulla). Double-labeling (sections from cortex) was performed as described above with antibodies recognizing AQP2 andcalbindin, respectively. AQP2 labeling was visualized with goatanti-rabbit IgG conjugated to 10-nm colloidal gold particles, whereascalbindin labeling was visualized with goat anti-mouse IgG conjugatedto 5-nm colloidal gold particles. Grids were stained with uranylacetate for 10 min and with lead citrate for 5 s.
9 ^5 m( ~6 ^) u5 Q% r
Quantitation of AQP2 Immunogold Labeling! z9 G' P$ c  g2 @! Q$ @

Electron micrographs were taken on a Philips CM100 electronmicroscope covering the entire area of the cell. Pictures were taken ofCNT cells and CCD principal cells from normal rats and CCD principalcells from rats treated with SR-121463B for 1 h. Gold particlesassociated with the apical and basolateral plasma membrane, as well asintracellular gold particles, were counted. The total labeling densitywas calculated as the total number of gold particles per cytoplasm areaexcluding the nucleus. We determined the total area of the cytoplasmusing a lattice square test system with a square size of 10 × 10 mm. The labeling of the basolateral plasma membrane as a fraction oftotal cell labeling was also determined in CCD cells from untreated andlong-term dDAVP-treated Brattleboro rats.: X; O* y# P& j% H8 o' W: H% D

RESULTS2 ~& j& s' s  D. \$ d9 C
. V! R' f1 v/ g
Basolateral AQP2 Localization in Normal Rats
! @" T" }2 I8 H9 S' s7 i' }! P
To investigate the basolateral distribution of AQP2 along the CNTand the collecting duct in normal rats, we performedimmunocytochemistry at the light and electron microscopic level. Figure 1 shows the localization of AQP2 in the different kidney zones [cortex, outer andinner stripe of outer medulla (ISOM), and inner medulla] in normalrats (Fig. 1, A-E, K, and L ).Immunoperoxidase microscopy revealed labeling of apical plasma membranedomains in all levels of the collecting duct in normal rats (Fig. 1, A-E ). In the cortex, AQP2 was localized either in bothapical and basolateral plasma membrane domains or mainly in apicalplasma membrane domains (Fig. 1, A and B ). Thesetwo strikingly different labeling patterns suggested a differentlocalization of AQP2 in CNT and CCD. To distinguish between CNT andcollecting ducts, we performed double immunolabeling for AQP2 andcalbindin (Fig. 1, K-N ). In the rat, calbindin is knownto be abundantly expressed in intracellular domains and in the nucleusin cells of the DCT and CNT, where it is involved in calciumreabsorption. Calbindin is reported only to be present in a fewcollecting duct cells in the superficial cortex in the rat ( 1, 28 ). Thus we identified tubules staining positive for calbindinas CNT, whereas calbindin negative-staining tubules were identified asCCD. This double-labeling procedure revealed that AQP2 is localized inboth apical and basolateral plasma membrane domains in CNT (Fig. 1 K ), whereas AQP2 is mainly situated in apical plasmamembrane domains in CCD principal cells (Fig. 1 L ). Doubleimmunoelectron microscopy using antibodies recognizing AQP2 andcalbindin, respectively, confirmed this labeling pattern. In CNT cells,AQP2 labeling was present in both the apical and basolateral plasmamembrane as well as intracellular vesicles and multivesicular bodies(Fig. 2, A and B ).In CCD principal cells (calbindin negative), very little labeling wasobserved in the basal part of the cell and the basolateral plasmamembrane (Fig. 3, A and B ). AQP2 was present mainly inthe apical part of the cell i.e., in the apical plasma membrane and inintracellular vesicles (Fig. 3, A and B ). Thedifferences in labeling of the basolateral plasma membrane in CNT andCCD were additionally confirmed by quantitation of the immunogoldlabeling in CNT cells and CCD principal cells (Table 1 ). The basolateral plasma membrane AQP2 labeling, determined as a fraction of total AQP2 cell labeling, wasonly 0.07 ± 0.02 ( n = 4) in the CCD, whereas inthe CNT the fraction was 0.33 ± 0.09 ( n = 3). Thetotal labeling density of AQP2 was lower in CNT compared with CCD(5.1 ± 1.7 vs. 23.8 ± 10.9 particles/µm 2 ).Calbindin is reported to be present in the CNT, but the specific celltypes expressing calbindin have not been reported. In the rat, theCNT consists of DCT cells, intercalated cells, CNT cells, and principalcells similar to those in the CCD ( 15 ). Therefore, wecannot conclude that all cells in the CNT containing both AQP2 andcalbindin are true CNT cells and not principal cells present in theCNT. However, we observed calbindin/AQP2-positive cells that had theappearance of true CNT cells (intermediate between CCD principal cellsand DCT cells), i.e., with deeper infoldings of the basal membranecompared with principal cells and with less mitochondria than the DCTcells ( 14 ). We also observed the presence of AQP2 in cells(both calbindin positive and negative) with an ultrastructure otherthan that of the CNT cells and the principal cells characterized by alightly stained cytoplasm, i.e., staining with heavy metals in theuranyl acetate and lead citrate used for counterstaining at theelectron microscopic level. These principal cells exhibitedlower cell height and had a more darkly stained cytoplasm, and theinfoldings of the basal membrane appeared less pronounced (Fig. 3 A ).
" Y" ?/ `5 Q, U
Fig. 1. Subcellular immunolocalization of aquaporin (AQP)-2 inwhole kidney sections from normal rats ( A-E, K, and L ) and rats treated with theV 2 -receptor antagonist SR-121463B (SR) for 60 min( F-J, M, and N ). A-J : paraffin-embedded sections were incubated withaffinity-purified anti-AQP2, and labeling was visualized withperoxidase-conjugated secondary antibody. A-E : AQP2localization in different kidney zones in a normal rat. OSOM, outerstripe of the outer medulla; ISOM, inner stripe of the outer medulla;IM, inner medulla. In the cortex, AQP2 is localized either both inapical and basolateral plasma membrane (BLM) domains (arrows, A ) or mainly in apical plasma membrane (APM) domains(arrows, B ). In collecting duct principal cells in OSOM andISOM, AQP2 is mainly localized in APM domains (arrows, C and D ). In inner medullary collecting duct (IMCD) principalcells, AQP2 labeling is seen in both APM and BLM domains as well asintracellular domains (arrowheads, E ). F-J :AQP2 localization in different kidney zones in a SR-121463B-treatedrat. SR-121463B treatment causes retrieval of AQP2 from APM domains(arrows) to intracellular domains (arrowheads) in all zones. K-N : immunofluorescent localization of AQP2 inconnecting tubules (CNT) and cortical collecting ducts (CCD) in normalrats ( K and L ) and in rats treated withSR-121463B for 60 min ( M and N ). Whole kidneysections were double labeled with polyclonal anti-AQP2 and monoclonalanti-calbindin D-28K to distinguish between CNT and CCD. Labeling wasvisualized with Alexa 488 anti-rabbit (green) and Alexa 546 anti-mouse(red), respectively. The presence of both calbindin (red) and AQP2(green) shows that the tubule is a CNT, whereas a tubule labeled withonly AQP2 (green) is a CCD. K : in CNT, AQP2 is localized inboth APM and BLM domains (arrows). L : in CCD, AQP2 islocalized mainly in APM domains. M and N :SR-121463B treatment causes retrieval of AQP2 from APM domains in bothCNT and CCD. Increased labeling of BLM domains is seen in CCD principalcells (arrows, N ). Magnification: ×600 in A-D and F-I; ×500 in E, J, L, and N; ×700 in K and M.' \( d$ Y% s8 |% q
  W0 W( \# s* b( S3 {
Fig. 2. Electron micrographs of ultrathin HM20 Lowicrylsections from kidney cortex (CNT) of normal rats ( A and B ) and rats treated with SR-121463B for 60 min ( C and D ). Sections were double immunogold labeled for AQP2(large arrows, large gold particles) and calbindin (small arrows, smallgold particles) to distinguish between CNT and CCD. The presence ofcalbindin as well as the cell ultrastructure indicates that the cellsare CNT cells. A : in the apical part of CNT cells in normalrats AQP2 labeling is associated with the apical plasma membrane (largearrows), intracellular vesicles (arrowheads), and multivesicularbodies (MVB). B : AQP2 labeling is also seen in the basalpart of the cells including the BLM. C : SR-121463B treatmentcauses retrieval of AQP2 from the apical plasma membrane tointracellular vesicles (arrowheads) present in the subapical part ofthe cell. D : there are no major changes in the labeling ofthe BLM after SR-121463B treatment (large arrows). L, lumen;M, mitochondrion. Magnification: ×63,000 in A; ×46,500 in B-D.

Fig. 3. Electron micrographs of ultrathin HM20 Lowicryl sectionsfrom kidney cortex (CCD) of normal rats ( A and B ). To distinguish between CNT and CCD, we double-immunogoldlabeled sections for AQP2 and calbindin. The absence of calbindin showsthat the cells are CCD principal cells. AQP2 is mainly present in theapical part of the cell in the apical plasma membrane (arrows) and inintracellular vesicles (arrowheads). Little labeling is present in thebasal part of the cell including the BLM (arrows). Besides principalcells with light staining of cytoplasm, i.e., staining with heavymetals in the uranyl acetate and lead citrate used for counterstainingat the electron microscopic level ( B ), we also observedprincipal cells that exhibited lower cell height and had a more darklystained cytoplasm and less pronounced infoldings of the basal membrane( A ). PC, principal cell; IC, intercalated cell.Magnification: ×46.500 in A, ×34,500 in B.
5 V% {8 ~+ {, X; h# N- F1 m- ^* j5 S
Table 1. Quantitation of immunogold labeling of AQP2 in CCD and CNT) \/ ~% W4 @" K" C/ a2 h% c  \
2 P2 O# p- s8 _! ]0 b4 i# f
Immunoperoxidase microscopy revealed that AQP2 is mainly localized inapical plasma membrane domains in principal cells in the inner (IMCD)and outer stripe of the outer medullary collecting duct (OMCD) (Fig. 1, C and D ). In IMCD principal cells, labeling wasseen of both apical and basolateral plasma membrane domains as well asof intracellular domains (Fig. 1 E ), consistent with previousfindings ( 22 ). Immunoelectron microscopy confirmed thepresence of AQP2 in the apical plasma membrane, intracellular vesicles,and the basolateral plasma membrane (Fig. 5 A ). Thus in CNTcells and in IMCD principal cells, AQP2 is present in both the apicaland basolateral part of the cell, including the basolateral plasmamembrane, whereas in OMCD and CCD principal cells AQP2 is mainlypresent in the apical part of the cell including the apical plasma membrane.% p& W0 a8 R* a4 ?$ Z9 w7 A/ w

Effects of 60-Min V 2 -Receptor Antagonist Treatment onBasolateral AQP2 Localization in Normal Rats

To investigate the effect of short-term V 2 -receptorantagonist treatment on AQP2 distribution in the basolateral part ofthe cell, we treated normal rats with SR-121463B for 60 min. Urine output was significantly increased in SR-121463B-treated rats (5.1 ± 0.5 vs. 0.51 ± 0.15 ml/h, n = 8 in eachgroup). V 2 -receptor antagonist treatment caused retrievalof AQP2 from apical plasma membrane domains to intracellular domains inall parts of the collecting duct including CNT, CCD, and OMCD as shownby light microscopy (Fig. 1, F-J, M, and N ). Observations in IMCD are consistent with previousresults showing retrieval of AQP2 from the apical plasma membrane inIMCD principal cells after 60 min of treatment with theV 2 -receptor antagonist OPC-31260 ( 4 ). Labelingof basolateral plasma membrane domains was seen in the CNT and all parts of the collecting duct after SR-121463B treatment (Fig. 1, F-J, M, and N ). Immunoelectronmicroscopy of ultrathin Lowicryl sections from the kidney cortex showedretrieval of AQP2 from the apical plasma membrane to intracellularvesicles present in the subapical part of the cell in CNT (Fig. 2 C ). There were no major changes in labeling of thebasolateral plasma membrane in response to SR-121463B treatment (Fig. 2 D ). In CCD principal cells, SR-121463B treatment causedretrieval of AQP2 from the apical plasma membrane (Fig. 4 ). The apical plasma membrane labeling, determined as a fraction of total labeling, decreased from 0.20 ± 0.04 in controls ( n = 4) to 0.006 ± 0.005 inSR-121463B-treated rats ( n = 4; Table 1 ). In contrast,there was increased labeling of the basal part of the cell includingthe basolateral plasma membrane (Fig. 4 ). This was also furtherconfirmed by quantitation of immunogold labeling in thebasolateral plasma membrane, which showed a major increase in thefraction of total labeling in the basolateral plasma membrane(0.29 ± 0.09 vs. 0.07 ± 0.02 in controls, n = 4). The total labeling density was unchangedafter SR-121463B treatment (21.0 ± 7.0 vs. 23.8 ± 10.9 incontrols, n = 4). In IMCD, there were no changes in thelabeling density of the basolateral plasma membrane (Fig. 5 B ). Thus short-termSR-121463B treatment of normal rats causes a major increase in AQP2labeling of the basolateral plasma membrane in CCD, whereas labeling ofthe basolateral plasma membrane in CNT and IMCD is unchanged.

Fig. 4. Electron micrographs of ultrathin HM20 Lowicryl sections fromkidney cortex (CCD) of rats treated with SR-121463B for 60 min.Treatment with SR-121463B causes retrieval of AQP2 from the apicalplasma membrane to intracellular vesicles (arrowheads) and the BLM(arrows). M, mitochondrion. Magnification, ×34,500.
1 l- {4 }& {& d. h( o
Fig. 5. Electron micrographs of ultrathin HM20 Lowicryl sections fromkidney inner medulla of a normal rat ( A ) and a rat treatedwith SR-121463B for 60 min ( B ). Sections were immunogoldlabeled for AQP2. A : AQP2 is present in the BLM in normalrats (arrows). B : SR-121463B treatment causes no changes inAQP2 labeling of the BLM. Magnification, ×34,500.

Basolateral AQP2 Localization in Kidneys from Brattleboro Rats

We investigated the subcellular localization of AQP2 invasopressin-deficient Brattleboro rats by immunoperoxidase andimmunofluorescent microscopy, in a fashion similar to the studiesperformed using normal rats (Figs. 6, A-D, and 7, A-E ). The effect ofshort- and long-term dDAVP treatment willbe described later (Figs. 6, E-H, and 7, F-O ). Immunoperoxidase microscopy of sections of whole kidney revealed two different patterns of labeling in the cortex. Insome segments, AQP2 was mainly localized apically (Fig. 6 B ), whereas in other segments AQP2 was mainly localized in the basolateral part of the cell (Fig. 6 A ). Double-labeling experiments withantibodies recognizing AQP2 and calbindin showed that cells of the CNTcontained AQP2 labeling of basolateral plasma membrane domains (Fig. 6 C ), whereas CCD principal cells mainly contained apicalAQP2 labeling (Fig. 6 D ). In general, the overall expressionof AQP2 was lower in CNT compared with CCD. Double-labeling experimentsat the immunoelectron microscopic level of ultrathin Lowicryl sectionsfrom the kidney cortex confirmed labeling of the basolateral plasmamembrane in CNT cells (Figs. 8 A and 10 A ). In CCD principal cells,little labeling was observed in the basal part of the cell includingthe basolateral plasma membrane (Figs. 9 A and 10 C ). In the inner and outerstripe of the outer medulla, AQP2 was mainly localized in apical plasma membrane domains (Fig. 7, A and B ), in IMCDprincipal cells, AQP2 was mainly localized in basolateral plasmamembrane domains with the basolateral labeling being more intense inthe inner third of IMCD (IM-3) compared with the outer third of IMCD(IM-1) (Fig. 7, C-E ). Immunoelectron microscopyconfirmed the subcellular distribution of AQP2 in sections from theouter third of IMCD. The labeling was mainly confined to thebasolateral part of the cell, including the basolateral plasma membrane(Figs. 9 C and 11 A ). Thus in untreated Brattlebororats, AQP2 labeling of the basolateral plasma membrane is mainlyobserved in CNT cells and in IMCD principal cells, very similar to thelabeling observed in normal rats.
4 D- f  J  S2 _
Fig. 6. Immunolocalization of AQP2 in the cortical part of wholekidney sections from control Brattleboro rats ( A-D ) andBrattleboro rats treated withdesamino-Cys 1, D -Arg 8 vasopressin(dDAVP) for 2 h ( E and F ) or 6 days( G and H ). Sections were incubated withaffinity-purified anti-AQP2, and labeling was visualized withperoxidase-conjugated secondary antibody ( A and B ). To distinguish between CNT and CCD, we incubated wholekidney sections with polyclonal anti-AQP2 and monoclonal anti-calbindinD-28K ( C-H ). Labeling was visualized with Alexa 488 anti-rabbit (green) and Alexa 546 anti-mouse (red), respectively. A and B : in the cortex of untreated Brattlebororats (controls from protocol 2 ) AQP2 is localized eithermainly in BLM domains (arrows, A ) or mainly in APM domains(arrows, B ). C and D : double labeling(controls from protocol 3 ) with AQP2 (green) and calbindin(red) revealed that AQP2 is mainly localized in BLM domains in cells ofthe CNT (arrows, C ) and mainly in APM domains in CCDprincipal cells (arrows, D ). E and F :labeling of APM domains is seen both in CNT and CCD after 2-h dDAVPtreatment (arrows, E ), and labeling of BLM domains ismaintained in CNT after dDAVP treatment (arrows, E ). G and H : labeling of both APM and BLM domains isseen in CNT and CCD after 6 days of dDAVP treatment (arrows).Magnification: ×500, A-C and E; ×300 D, F, and H; ×700 G.

Fig. 7. Immunolocalization of AQP2 in the medullary collecting duct ofcontrol Brattleboro rats (controls from protocol 2 )( A-E ) and Brattleboro rats treated with dDAVP for2 h ( F-J ) or 6 days ( K-O ). A and B : in OSOM and ISOM of untreatedBrattleboro rats AQP2 is localized mainly in APM domains in collectingduct principal cells (arrows). C-E : AQP2 is mainlylocalized in BLM domains in IMCD principal cells with the basolaterallabeling being more intense in the inner third of IMCD (IM-3) comparedwith the outer third of IMCD (IM-1, arrows). F-J : after2 h of dDAVP exposure, AQP2 was localized in APM domains in allparts of the medullary collecting duct (arrows). Decreased labeling ofBLM domains was seen in IMCD principal cells ( H-J ). K-O : after 6 days of dDAVP exposure, AQP2 was localizedin APM domains in all parts of the medullary collecting duct (arrows).In OSOM principal cells, increased labeling of BLM domains was seen(arrow, K ). Decreased labeling of BLM domains was seen inIMCD principal cells ( M-O ). Magnification,×700.
! S% Z6 N4 ]8 ~+ ~! B  O
Fig. 8. Electron micrographs of ultrathin HM20 Lowicryl sections from thekidney cortex (CNT) of an untreated Brattleboro rat ( A ) anda Brattleboro rat treated with dDAVP for 2 h ( B ).Sections were double immunogold labeled for AQP2 (large arrows, largegold particles) and calbindin (small arrows, small gold particles) todistinguish between CNT and CCD. The presence of calbindin as well asthe cell ultrastructure indicates that the cells are CNT cells. A : AQP2 is present in the basal part of the cell includingthe BLM in an untreated Brattleboro rat (large arrows). B :2 h after dDAVP injection there is no major changes in AQP2labeling of the BLM (large arrows). Magnification, ×34,500.
( t8 k9 }8 t) F4 q4 [
Fig. 9. Electron micrographs of ultrathin HM20 Lowicryl sectionsfrom kidney cortex (CCD, A and B ) and innermedulla ( C and D ) of untreated Brattleboro rats( A and C ) and Brattleboro rats treated with dDAVPfor 2 h ( B and D). A and B :sections from cortex were double-immunogold labeled for AQP2 andcalbindin to distinguish between CNT and CCD. The absence of calbindinshows that the cells are CCD principal cells. A : in CCD ofan untreated Brattleboro rat little labeling of the BLM is seen(arrows). B : 2 h after dDAVP injection there is nomajor changes in AQP2 labeling of the BLM (arrows). C : inthe inner medulla, AQP2 is present in the basal part of the cellincluding the BLM in an untreated Brattleboro rat (arrows). D : 2 h after dDAVP injection, there is a slightdecrease in AQP2 labeling of the BLM (arrows). Magnification,×34,500.

Fig. 10. Electron micrographs of ultrathin HM20 Lowicryl sectionsfrom the kidney cortex of untreated Brattleboro rats ( A and C ) and Brattleboro rats treated with dDAVP for 6 days( B and D ). Sections were double-immunogoldlabeled for AQP2 (large arrows, large gold particles) and calbindin(small arrows, small gold particles) to distinguish between CNT andCCD. The presence of calbindin as well as the cell ultrastructureindicates that the cells are CNT cells ( A and B ). A : in CNT of an untreated Brattleboro rat AQP2 is present inthe BLM (large arrows). B : AQP2 labeling of the BLM ismoderately increased after 6 days' dDAVP treatment in CNT cells (largearrows). C : in CCD of an untreated Brattleboro rat littlelabeling of the BLM is seen (large arrows). D : 6 days ofdDAVP treatment causes a major increase in AQP2 labeling of the basalpart of the cell including the BLM in CCD principal cells (largearrows). The absence of calbindin shows that the cells are CCDprincipal cells. Magnification, ×34,500." |( j$ A* z( W0 m5 p6 H% @2 a

Fig. 11. Electron micrographs of ultrathin HM20 Lowicryl sections fromkidney inner medulla of an untreated Brattleboro rat ( A ) anda Brattleboro rat treated with dDAVP for 6 days ( B ). A : AQP2 is present in the basal part of the cell includingthe BLM in an untreated Brattleboro rat (arrows). B : thereis a slight increase in AQP2 labeling of the BLM after 6 days of dDAVPtreatment (arrows). MVB, multivesicular body. Magnification,×34,500.+ d' P$ V; H+ _) W* A

Effects of 2-h dDAVP Treatment on Basolateral AQP2Localization in Brattleboro Rats

Two hours of dDAVP treatment of Brattleboro rats caused asignificant decrease in urine output (1.9 ± 0.4 vs. 20 ± 3.3 ml/2 h, n = 13). Immunocytochemistry showed thatshort-term dDAVP treatment induces a marked increase in labeling ofapical plasma membrane domains in cells of the CNT (Fig. 6 E )and in IMCD principal cells (Fig. 7, H-J ). In contrast,in IMCD principal cells, AQP2 labeling of the basolateral plasmamembrane domains was decreased after short-term dDAVP treatment (Fig. 7, H-J ). Immunoelectron microscopy showed nosignificant changes in the labeling of the basolateral plasma membranein CNT cells and in CCD principal cells (Figs. 8 B and 9 B ), whereas in the IMCD the AQP2 immunogold labelingdensity of the basolateral plasma membrane was slightly decreased (Fig. 9 D ). Thus after 2 h of acute dDAVP treatment there isno increase in the basolateral plasma membrane expression of AQP2.: }* R! u" W1 w/ [
5 P' @3 J8 z3 w- _* L0 d' h
Effects of 6-Day dDAVP Treatment on Basolateral AQP2Localization in Brattleboro Rats

Six days of dDAVP treatment caused a dramatic decrease in urineoutput (12 ± 1.4 vs. 219 ± 20 ml/24 h, n = 9, the last 24 h of treatment).Immunocytochemistry showed that long-term dDAVP treatment ofBrattleboro rats induced an increase in apical plasma membrane domainlabeling in all parts of the collecting duct (Figs. 6, G and H, and 7, K-O ). Labeling of basolateralplasma membrane domains was maintained in CNT cells (Fig. 6 G ) and increased in CCD and OSOM principal cells (Figs. 6 H and 7 K ). The basolateral plasma membranedomain labeling was decreased in IMCD principal cells (Fig. 7, M-O ). Immunoelectron microscopy of sections of kidneycortex showed that the labeling of the basolateral plasma membrane wasmoderately increased in CNT cells (Fig. 10 B ). There was a majorincrease in labeling of the basal part of the cell, including thebasolateral plasma membrane in CCD principal cells (Fig. 10 D ). Quantitation of immunogold labeling showed that the labeling of the basolateral plasma membrane taken as a fraction oftotal cell labeling was increased in dDAVP-treated rats (0.21 ± 0.04 vs. 0.05 ± 0.01 in controls, n = 4). Thetotal labeling density in CCD was unchanged (8.8 ± 1.4 vs.8.1 ± 1.4 in controls, n = 4). After long-termdDAVP treatment, AQP2/calbindin-positive tubules were seen with ahigher frequency compared with control Brattleboro rats. Some of thesecells had the appearance of CCD principal cells, suggesting that partof the CCD becomes calbindin positive after dDAVP exposure or thatthere may also be a longer region of the CNT that expresses AQP2 inresponse to long-term treatment with dDAVP. Consistent with this,previous studies have indicated that the CNT of Brattleboro rats has anupstream region lacking AQP2 and that chronic vasopressin treatmentinduces AQP2 expression in this region ( 5 ). In the outerthird of the IMCD, AQP2 labeling in the basal part of the cells wassignificantly decreased in response to long-term dDAVP treatment,whereas the labeling density of the basolateral plasma membrane wasslightly increased (Fig. 11 B ). The overall expressionof AQP2 also increased in IMCD. Thus long-term dDAVP treatment causes amajor increase in the basolateral plasma membrane expression of AQP2 inCCD principal cells, which was not associated with increased proteinexpression in CCD (i.e., determined as total AQP2 immunogold labeling).In contrast, CNT cells and IMCD principal cells exhibit only a moderate increase in basolateral plasma membrane labeling.1 c! ^$ i, E6 b" u. L
3 W4 ^: Q( O* g" h8 M
DISCUSSION
; n% H6 @6 r/ E. M; u; p0 ]
Immunocytochemistry at the light- and electron microscopicallevel revealed an axial heterogeneity in the basolateral localization of AQP2 along the CNT and the collecting duct segments in normal ratsand in vasopressin-deficient Brattleboro rats. AQP2 labeling of thebasolateral part of the cell, including the basolateral plasmamembrane, was seen mainly in CNT cells and IMCD principal cells,whereas CCD and OMCD principal cells exhibited little basolateral labeling. Short-term dDAVP treatment of Brattleboro rats caused no netincrease in the labeling of the basolateral plasma membrane in any segment, whereas long-term (6 days) dDAVP treatment of Brattleboro rats caused a major increase in basolateral plasma membraneAQP2 labeling in CCD principal cells, but not in IMCD cells. AcuteV 2 -receptor antagonist treatment of normal rats caused nodecrease in basolateral plasma membrane AQP2 labeling. Incontrast, increased AQP2 expression in the basolateral plasma membranewas observed in CCD in response to acute V 2 -receptorantagonist treatment.

Axial Heterogeneity in Basolateral Localization of AQP2  C; `" ~, i: p) \7 ~, F% N, l
- Y( ]2 e) `% ?+ M2 v4 J
Our results demonstrated that there is an axial heterogeneity inthe basolateral AQP2 localization. AQP2 labeling of the basolateral plasma membrane was most prominent in CNT cells and in IMCD principal cells in both normal rats and Brattleboro rats. AQP2 labeling of thebasolateral plasma membrane has been reported previously in IMCDprincipal cells in normal rats, as well as in a undefined corticalsegment i.e., collecting duct or CNT ( 22 ). Colemann et al.( 5 ) have recently reported that AQP2 occurs in basolateral plasma membrane domains of CNT cells to a variable extent, and Jeon etal. ( 10 ) have also shown labeling of basolateral plasma membrane domains in CNT cells. In this study, we provide evidence ofextensive AQP2 labeling in the basolateral plasma membrane usingquantitative immunoelectron microscopy.1 {9 Z* E( c! m3 h& r4 A
3 y6 i2 _6 q8 w5 b" S
The mechanisms involved in the targeting of AQP2 expression tothe basolateral plasma membrane are not well defined. Van Balkom et al.( 33 ) tested the hypothesis that the presence of AQP2 atthe basolateral plasma membrane may be due to heterotetramerization with AQP3 and AQP4. However, coexpression studies in oocytes indicated that this is not the case.
6 K/ f7 u* ?  }$ D, H
Hypertonicity appears to have an effect on the expression level of somerenal aquaporins. A hypertonicity-responsive element has beenidentified in the human AQP1 and AQP2 genes( 18, 30 ), and the expression of AQP3 has been shown toincrease in response to hypertonicity in MDCK cells ( 19 ).Recent in vitro studies by van Balkom et al. ( 33 ) havesuggested that hypertonicity is involved in basolateral targeting ofAQP2. Using different osmolytes, they showed that a gradual increase(over a 3-day period) in osmolarity of the surrounding medium causesbasolateral AQP2 targeting in response to forskolin stimulation inAQP2-transfected MDCK cells. This theoretically could explain thebasolateral expression pattern in IMCD in normal rats being in ahypertonic milieu (exposed to high basolateral tonicity). But thisprobably does not offer an explanation for the basolateral localizationof AQP2 in CNT, nor the difference in basolateral AQP2 expression inCNT and CCD. In contrast to the CNT, we observed little or no AQP2 inthe basolateral plasma membrane in CCD principal cells.Although the absence of basolateral AQP2 in CCD principalcells correlates with the hypothesis of an important role ofhypertonicity on basolateral AQP2 targeting, the extensive basolaterallabeling in CNT contradicts this hypothesis. The CNT and CCD are bothembedded in one to plasma isotonic interstitium, and the luminalosmolality is low, ~100 mosmol/kgH 2 O. Also in the kidneymedulla there are differences in basolateral membrane expression ofAQP2. In the medullary collecting duct, which is embedded in ahyperosmotic interstitium (the luminal osmolality ranging from 200 to3,000 mosmol/kgH 2 O in mammals), we observed significantAQP2 labeling of the basolateral plasma membrane in the IMCD. InBrattleboro rats the basolateral expression was more intense in theinner third of IMCD. In contrast, there was no significant basolateralAQP2 labeling in collecting duct principal cells in ISOM. Thus there isa progressive decrease in basolateral AQP2 expression going from theinitial IMCD to the OMCD with a change in the presence or absence ofAQP2 in the basolateral plasma membrane between the initial IMCD andthe terminal ISOM collecting duct. Although the interstitial osmolalityalso decreases through the medulla, there is no evidence for a similarmajor change in the interstitial osmolality in the same region, whichcould explain the significant differences in basolateral AQP2expression. Thus the hypertonicity theory most likely does not explainthe basolateral labeling in the CNT in normal rats nor the basolateralAQP2 labeling in Brattleboro rat CNT and IMCD principal cells, whichare in a more or less permanent state of water diuresis. Moreover,preliminary results revealed a maintained or even increased basolateralAQP2 labeling in IMCD principal cells in kidneys from Wistar ratstreated with the loop diuretic furosemide in vivo (Christensen andNielsen, unpublished observations). This also speaks againstthe hypothesis that hypertonicity induces basolateral AQP2 targeting inrat kidney.

The presence of basolateral plasma membrane labeling invasopressin-deficient Brattleboro rats indicate that factors other thanvasopressin are responsible for the expression of AQP2 at thebasolateral plasma membrane. There may, however, be anextrahypothalamic production of vasopressin or even a low hypothalamicproduction of vasopressin in Brattleboro rats ( 32 ), and itcould be speculated that even though these potentially very lowvasopressin levels have no physiological effects (i.e., antidiureticeffect), the levels may be high enough to cause basolateral AQP2targeting. This would be consistent with previous findings that theoverall expression of AQP2 in Brattleboro rats decreases in response to V 2 -receptor antagonist treatment ( 25 ).However, if low vasopressin levels are responsible for the basolateralplasma membrane labeling in untreated Brattleboro rats, it wouldexpected that acute vasopressin treatment would increase basolateralplasma membrane labeling, which is not the case. Oxytocin is anotherhormone with antidiuretic effect ( 3 ), and treatment ofnormal rats with this hormone has been reported to induce both apicaland basolateral targeting of AQP2, an effect that was abolished bypretreatment with a V 2 -receptor antagonist( 10 ). Thus oxytocin may be involved in basolateral AQP2targeting in Brattleboro rats.

Acute V 2 -Receptor Agonist Treatment Causes No Increasein Basolateral AQP2 Labeling
- L! K; [  h' D( i! J9 u7 D
Our data show no evidence of basolateral targeting of AQP2 inresponse to 2-h dDAVP treatment in Brattleboro rats. This was the casein all AQP2-containing segments in the kidney tubule. Recentnonphysiological in vitro studies have shown that stimulation ofdissected kidney slices (from normal rats) with vasopressin andforskolin for 15 min causes translocation of AQP2 from cytoplasmic vesicles to basolateral plasma membrane domains in IMCD principal cells, with the most prominent basolateral staining in the distal partof IMCD ( 33 ). However, when the experiment was performed with kidney slices from Brattleboro rats, there was mainly staining ofapical plasma membrane domains after 15 min of forskolin/vasopressin treatment, consistent with our results ( 33 ). It could beargued that treatment of normal rats with dDAVP for a shorter period than 2 h, e.g., minutes, would cause basolateral targeting of AQP2in vivo. However, previous studies have indicated that this is not thecase. Treatment of normal rats with dDAVP for 20 min caused no increasein the labeling density of the basolateral plasma membrane in IMCDprincipal cells ( 17 ), and there was no change in thelabeling of the basolateral plasma membrane in IMCD principal cells ofBrattleboro rats after 15 min of vasopressin stimulation( 34 ). Furthermore, it has recently been shown that, after30 min of vasopressin treatment, there is no increase in AQP2 labelingof basolateral plasma membrane domains in any regions of the kidney( 10 ).  t* f' t: U; B5 R* ], }0 M" s5 e, A

Long-Term V 2 -receptor Agonist Treatment IncreasesBasolateral AQP2 Labeling

AQP2 labeling of the basolateral plasma membrane was onlyprominent in IMCD principal cells and CNT cells in untreatedBrattleboro rats, and long-term dDAVP treatment caused only minorincreases in basolateral plasma membrane AQP2 labeling. In contrast,little AQP2 labeling of the basolateral plasma membrane was observed inCCD principal cells in untreated Brattleboro rats, and in this segmenta major increase in AQP2 expression in the basolateral plasma membranewas observed after long-term dDAVP treatment. Quantitation of goldparticles in CCD principal cells showed a significantly increased AQP2labeling in the basolateral plasma membrane (expressed as a fraction oftotal cell labeling) after long-term dDAVP treatment. Thus thisdemonstrates that the increase in basolateral AQP2 labeling in the CCDwas not a consequence of increased AQP2 protein expression aspreviously demonstrated ( 7, 29 ). Increased AQP3 proteinexpression has been shown in response to long-term dDAVP treatment ofBrattleboro rats ( 29 ). AQP3 is present mainly in thebasolateral plasma membrane, and AQP3 trafficking is not regulated byacute vasopressin ( 8 ). Only a small fraction of AQP3 islocated in intracellular vesicles, most likely representing newlysynthesized AQP3 protein in transit from the Golgi to the basolateralplasma membrane ( 8 ). Thus similar to AQP3, AQP2 may beconstitutively expressed at the basolateral plasma membrane andparticipate in water transport across the basolateral membrane,although the contribution of AQP2 as an exit pathway is likely quitesmall, since AQP3 knockout mice are severely polyuric( 13 ). However, the fact that AQP3 knockout mice do survivein contrast to AQP2 and vasopressin V 2 -receptor knockoutmice may be related to a role of AQP4 and, potentially, also a role ofAQP2 in the basolateral plasma membranes.0 ^) V2 _# B% N1 u+ c( a( R& j

Changes in the Subcellular Localization of AQP2 in Response toAcute V 2 -Receptor Antagonist Treatment* M/ H; L+ z- x, z7 z7 K

Immunoelectron microscopy revealed that V 2 -receptorantagonist treatment causes increased labeling of AQP2 in thebasolateral plasma membrane of CCD principal cells, whereas thebasolateral labeling density is unchanged in CNT and IMCD.Interestingly, the effect of V 2 -receptor antagonisttreatment on the subcellular localization of AQP2 differed considerablyin CCD and CNT. Although AQP2 was internalized from the apical plasmamembrane in both CNT and CCD, AQP2 was retrieved to intracellularvesicles in the subapical part of the CNT cell with unchanged labelingof the basolateral plasma membrane, whereas in the CCD AQP2 wasretrieved to the basal part of the cell with increased labeling of thebasolateral plasma membrane. The fact that we did not observe aretrieval of AQP2 from the basolateral plasma membrane stronglysupports the view that AQP2 is not targeted to the basolateral plasmamembrane by short-term V 2 -receptor stimulation (vasopressinor dDAVP).  r" B; }% W$ N

Basolateral Localization of AQP2" C3 ^* T5 w- l3 G
# N% q/ `. T  w- u) D. l0 d: K
The results demonstrate that there is a significant basolateralAQP2 targeting in connecting tubule cells and in IMCD cells. There isno change in basolateral targeting in response to acute vasopressintreatment of vasopressin-deficient Brattleboro rats, consistent withprevious studies ( 10, 17, 34 ). This indicates thatbasolateral targeting is not vasopressin mediated, which is consistentwith the observation that significant basolateral AQP2 targeting isselectively seen in the CNT and IMCD but not or to a much lesser extentin the CCD and OMCD. This selective basolateral targeting also stronglyspeaks against an effect of interstitial osmolality, since it isunlikely that there are major differences in interstitial osmolalityassociated with the CNT and CCD. Thus the signaling mechanisms remainuncertain, but the axial heterogeneity strongly suggests thatcell-specific mechanisms are involved. It should be emphasized thatthere, indeed, are marked ultrastructural differences between, e.g.,CNT cells, collecting duct principal cells in the CCD, and IMCD cells.Moreover, there are marked differences in the expression pattern oftransporters, e.g., the vasopressin-regulated urea transporter, whichis expressed only in the terminal IMCD, not in principals cells in moreproximal parts of the collecting duct (for a review, see Ref. 27 ). Also, AQP4 is expressed to a greater extent in theIMCD than in CCD. Thus cell-specific mechanisms are likely to beinvolved in basolateral AQP2 expression. The targeting mechanisms aremost likely specific to AQP2 since epithelial Na channelsubunits are expressed only in the apical plasma membrane and invesicles but not in the basolateral plasma membrane ( 9, 12 ). Additional studies are necessary to establish thesignaling and potential regulation of basolateral AQP2.
! s+ m8 k! c! |2 h  `
Summary
5 G5 }- y; H' ^1 [6 l: i7 S2 Q! o
Our results show that there is a marked axial heterogeneity in theexpression levels of AQP2 in the basolateral plasma membrane along theCNT and collecting duct subsegments. AQP2 expression in the basolateralplasma membrane was not increased by short-term dDAVP treatment inBrattleboro rats. Moreover, acute V 2 -receptor antagonisttreatment of normal rats did not cause a decrease in basolateral plasmamembrane labeling. Long-term (6 days) dDAVP treatment of Brattlebororats caused major increases in basolateral plasma membrane labeling ofAQP2 in CCD principal cells and moderate increases in CNT cells andIMCD principal cells.) R$ A% A% t* Y) g% [

ACKNOWLEDGEMENTS% l2 V% A2 ~- ~: n4 S7 G* k
4 J$ \' m; @" f
We thank Zhila Nikrozi, Inger Merete S. Paulsen, Merete Pedersen,Ida Maria Jalk, and Gitte Kall for expert technical assistance. We alsothank Claudine Serradeil-Le Gal, Sanofi-Synthélabo Recherche, France, for the kind gift of the SR121463B compound.
      【参考文献】
1. Bindels, RJ,Hartog A,Timmermans JA,andvan Os CH. Immunocytochemical localization of calbindin-D28k, calbindin-D9k and parvalbumin in rat kidney. Contrib Nephrol 91:7-13,1991 .

% j2 V# L( K# M3 ]& @! r5 ^( I- P
- c/ |, V" s2 `/ k* M+ e$ [4 K
2. Bouley, R,Breton S,McLaughlin M,andBrown D. Vasopressin (VP) stimulates apical and basolateral membrane insertion, and clathrin coated pit localization of AQP2 in collecting duct principal cells (Abstract). J Am Soc Nephrol 11:14A,2000.. u! Q! S4 }" @
7 p$ {- l  V* }- r

% d, j! g% w$ z0 m7 {0 w1 [
3. Chou, CL,DiGiovanni SR,Mejia R,Nielsen S,andKnepper MA. Oxytocin as an antidiuretic hormone. I. Concentration dependence of action. Am J Physiol Renal Fluid Electrolyte Physiol 269:F70-F77,1995 .: k- M, G2 }7 C$ x3 ^5 {8 \  ]

! f+ {8 \- J9 `" V
: k! f; d4 N. j
4. Christensen, BM,Marples D,Jensen UB,Frøkiær J,Sheikh Hamad D,Knepper M,andNielsen S. Acute effects of vasopressin V 2 -receptor antagonist on kidney AQP2 expression and subcellular distribution. Am J Physiol Renal Physiol 275:F285-F297,1998 .- T1 N7 a0 C  n. G& A3 \( ~4 k
" k9 S7 H  s/ F& H7 Q

5. Coleman, RA,Wu DC,Liu J,andWade JB. Expression of aquaporins in the renal connecting tubule. Am J Physiol Renal Physiol 279:F874-F883,2000 .6 ]5 x# I% w- D. [4 V" l' s
! Z, `7 q4 v( Y% u) J5 `9 X  _* P

6. Deen, PM,Rijss JP,Mulders SM,Errington RJ,van Baal J,andvan Os CH. Aquaporin-2 transfection of Madin-Darby canine kidney cells reconstitutes vasopressin-regulated transcellular osmotic water transport. J Am Soc Nephrol 8:1493-1501,1997 .$ ?' `& M: x  B2 x

3 K1 Y" f* F- J: O% ~
7. DiGiovanni, SR,Nielsen S,Christensen EI,andKnepper MA. Regulation of collecting duct water channel expression by vasopressin in Brattleboro rat. Proc Natl Acad Sci USA 91:8984-8988,1994 .; d- O0 Y2 I. H' O4 \4 q' o

) r3 U0 m0 \8 C6 k2 H" @
- D0 L8 I4 `! E
8. Ecelbarger, CA,Terris J,Frindt G,Echevarria M,Marples D,Nielsen S,andKnepper MA. Aquaporin-3 water channel localization and regulation in rat kidney. Am J Physiol Renal Fluid Electrolyte Physiol 269:F663-F672,1995 .

9. Hager, H,Kwon TH,Vinnikova AK,Masilamani S,Brooks HL,Frøkiær J,Knepper MA,andNielsen S. Immunocytochemical and immunoelectron microscopic localization of -, -, and -ENaC in rat kidney. Am J Physiol Renal Physiol 280:F1093-F1106,2001 .% R" Y% j2 ?. L1 A/ [# y# G7 }
, Q9 _- n9 s8 h" l/ N5 J* I

10. Jeon, US,Joo KW,Na KY,Kim YS,Lee JS,Kim J,Kim GH,Nielsen S,Knepper MA,andHan JS. Oxytocin induces apical and basolateral redistribution of aquaporin-2 in rat kidney. Nephron 93:E36-E45,2003.
( _. |' I* }: p) g6 s& }& c

11. Katsura, T,Verbavatz JM,Farinas J,Ma T,Ausiello DA,Verkman AS,andBrown D. Constitutive and regulated membrane expression of aquaporin 1 and aquaporin 2 water channels in stably transfected LLC-PK1 epithelial cells. Proc Natl Acad Sci USA 92:7212-7216,1995 .

12. Loffing, J,Pietri L,Aregger F,Bloch-Faure M,Ziegler U,Meneton P,Rossier BC,andKaissling B. Differential subcellular localization of ENaC subunits in mouse kidney in response to high- and low-Na diets. Am J Physiol Renal Physiol 279:F252-F258,2000 .* [. d5 {$ t4 y

( z' m% F8 j% S* d4 c  Y" h% ^: j" N
% I/ ?$ b' c5 ?* I' x: U# w! r
13. Ma, T,Song Y,Yang B,Gillespie A,Carlson EJ,Epstein CJ,andVerkman AS. Nephrogenic diabetes insipidus in mice lacking aquaporin-3 water channels. Proc Natl Acad Sci USA 97:4386-4391,2000 .7 l. Y5 P5 Q9 d- t
- J: h5 M* w1 r# _

14. Madsen, KM,andTisher CC. Structural-functional relationships along the distal nephron. Am J Physiol Renal Fluid Electrolyte Physiol 250:F1-F15,1986 .$ l) `# p% K( ~) ^7 [( w
3 p- v/ |9 P' m6 E: N; c5 T  v

" Q- R4 J0 j  p% @! ]2 W
15. Madsen, KM,Verlander JW,andTisher CC. Relationship between structure and function in distal tubule and collecting duct. J Electron Microsc Tech 9:187-208,1988  ./ s6 f5 q: u+ U% a( I9 h7 F

+ e5 K- u% L: c6 V2 p
16. Maric, K,Oksche A,andRosenthal W. Aquaporin-2 expression in primary cultured rat inner medullary collecting duct cells. Am J Physiol Renal Physiol 275:F796-F801,1998 .

5 m) H6 H. G2 f! ]) ^9 o9 D! d
17. Marples, D,Knepper MA,Christensen EI,andNielsen S. Redistribution of aquaporin-2 water channels induced by vasopressin in rat kidney inner medullary collecting duct. Am J Physiol Cell Physiol 269:C655-C664,1995 .5 D7 R; @! ?; e, c& o
2 k$ {  G6 a1 T5 ?$ }4 K+ z- x

18. Matsumura, Y,Uchida S,Sasaki S,andMarumo F. Identification of the hypertonicity-responsive element in human aquaporin-2 promoter (Abstract). J Am Soc Nephrol 6:326,1995.

* z2 Q0 C1 `+ n" o
, g  W! A/ G! b* [5 N' l; \
19. Matsuzaki, T,Suzuki T,andTakata K. Hypertonicity-induced expression of aquaporin 3 in MDCK cells. Am J Physiol Cell Physiol 281:C55-C63,2001 .
% b, t) B; U8 I8 b* N7 I
$ \4 z8 f% C  d# R) r1 r
, ^4 r0 X4 v4 t- @  u9 y& e
20. Maunsbach, AB. Embedding of cells and tissues for ultrastructural and immunocytochemical analysis.In: Cell Biology. A Laboratory Handbook, edited by Celis JE.. San Diego, CA: Academic, 1994, p. 117-125.! ?+ }4 X3 y6 m/ m, L
" }1 y! U8 s6 ?% Z

21. Nielsen, S,Chou CL,Marples D,Christensen EI,Kishore BK,andKnepper MA. Vasopressin increases water permeability of kidney collecting duct by inducing translocation of aquaporin-CD water channels to plasma membrane. Proc Natl Acad Sci USA 92:1013-1017,1995 .
+ I, p$ b0 a  f) A: S! W
4 O5 l, T9 G0 R% R! [$ d3 J4 d
/ L* Z8 ~" U. {6 t+ Z* B
22. Nielsen, S,DiGiovanni SR,Christensen EI,Knepper MA,andHarris HW. Cellular and subcellular immunolocalization of vasopressin-regulated water channel in rat kidney. Proc Natl Acad Sci USA 90:11663-11667,1993 .
# `! }! k+ v7 r9 r
/ R8 X3 b7 I2 b8 i  s4 Q" n
1 P: [$ ]; x6 L7 X, ?
23. Nielsen, S,Frøkiær J,Marples D,Kwon TH,Agre P,andKnepper MA. Aquaporins in the kidney: from molecules to medicine. Physiol Rev 82:205-244,2002 .

  ?) ]/ \6 [  P0 c+ N* ?  u

24. Nielsen, S,Pallone T,Smith BL,Christensen EI,Agre P,andMaunsbach AB. Aquaporin-1 water channels in short and long loop descending thin limbs and in descending vasa recta in rat kidney. Am J Physiol Renal Fluid Electrolyte Physiol 268:F1023-F1037,1995 .- x0 g% f) f1 |" t  S
' V8 V4 |8 _' o
( _1 }# @; x8 N  ^) d

25. Promeneur, D,Kwon TH,Frokiaer J,Knepper MA,andNielsen S. Vasopressin V 2 -receptor-dependent regulation of AQP2 expression in Brattleboro rats. Am J Physiol Renal Physiol 279:F370-F382,2000 .
. u) P1 m6 T6 d1 X9 o

: M6 ^1 ^- r8 j+ e
26. Sabolic, I,Katsura T,Verbavatz JM,andBrown D. The AQP2 water channel: effect of vasopressin treatment, microtubule disruption, and distribution in neonatal rats. J Membr Biol 143:165-175,1995  . [Corrigenda. J Membr Biol 145: May 1995, 107-108.]

! Y0 P7 S- F: S6 ^! j+ U2 i& b

27. Sands, JM. Molecular approaches to urea transporters. J Am Soc Nephrol 13:2795-2806,2002 .

' C+ `7 i; [* u& ~  w8 W
: t3 i$ t9 P+ V# O, r% n
28. Taylor, AN,McIntosh JE,andBourdeau JE. Immunocytochemical localization of vitamin D-dependent calcium-binding protein in renal tubules of rabbit, rat, and chick. Kidney Int 21:765-773,1982  .3 h0 L! T: ~4 [. }9 @

4 \) @$ m; s' M; ~" e: g' V
29. Terris, J,Ecelbarger CA,Nielsen S,andKnepper MA. Long-term regulation of four renal aquaporins in rats. Am J Physiol Renal Fluid Electrolyte Physiol 271:F414-F422,1996 .6 D' `$ r. F; i5 n
2 L' e  d9 l% z" o
! E' f* z: v& v/ \

30. Umenishi, F,andSchrier RW. Identification and characterization of a novel hypertonicity-responsive element in the human aquaporin-1 gene. Biochem Biophys Res Commun 292:771-775,2002  .! F2 |: @2 i" g1 a
9 r  @4 y$ h6 p; m+ z0 L9 C

31. Valenti, G,Frigeri A,Ronco PM,D'Ettorre C,andSvelto M. Expression and functional analysis of water channels in a stably AQP2-transfected human collecting duct cell line. J Biol Chem 271:24365-24370,1996 .
9 |5 |, N1 h# @. f7 I; B& d
9 ~4 h: n" r3 @- h0 `; H# c

32. Valtin, H. Genetic models of diabetes insipidus.In: Handbook of Physiology. Renal Physiology, edited by Windhager EE.. Bethesda, MD: Am. Physiol. Soc, 1992, sect. 8, vol. II, chapt. 28, p. 1281-1316.% t/ X: h2 g' z9 z# @
, {% ?+ p- A/ r$ H9 H9 W

2 C( K3 \9 w+ C4 ~, {+ |% m
33. Van Balkom BW, van Raak M, Breton S, Pastor-SolerN, Bouley R, Van Der Sluijs P, Brown D, and Deen PM. Hypertonicity is involved in redirecting the aquaporin-2 water channelinto the basolateral, instead of the apical, plasma membrane of renalepithelial cells. J Biol Chem 2002./ K4 B7 `/ Y3 ~* c

4 l) J0 D" S0 {# {" H% c- o% X

34. Yamamoto, T,Sasaki S,Fushimi K,Ishibashi K,Yaoita E,Kawasaki K,Marumo F,andKihara I. Vasopressin increases AQP-CD water channel in apical membrane of collecting duct cells in Brattleboro rats. Am J Physiol Cell Physiol 268:C1546-C1551,1995 .

收藏0 分享0 顶0 踩0