Mineralocorticoid regulation of epithelial Na channels is maintained in a mous - 《American Journal of Physiology》美国生理学杂志干细胞之家



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 ... 26 下一页

返回列表

查看: 575689\|回复: 250	go Mineralocorticoid regulation of epithelial Na channels is maintained in a mous [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 13:41 |只看该作者 |倒序浏览 |打印

作者：Anke Dahlmann, Sylvain Pradervand, Edith Hummler, Bernard C. Rossier, Gustavo Frindt, and Lawrence G. Palmer作者单位：1 Department Physiology of Biophysics, WeillMedical College of Cornell University, New York, New York 10021; and and 2 Institut de Pharmacologie et de Toxicologie,Université de Lausanne, 1005 Lausanne, Switzerland ; A0 D( c# U8 X% f/ ^

   * Q# f0 m" t$ ~+ E0 @, c5 y
      1 J" ?4 L# u# y  W
                     8 _1 Y3 D% |8 m% u& o
         ' x1 k* E- C) F1 x* t  W" X

         8 i, Q1 l. r5 A8 N  Z4 U/ v

                  * \" ~* l+ y3 }* W1 b6 h$ @5 e
        @( \5 Q: |& z4 `


      【摘要】
   Currents through epithelial Na channels (ENaCs) were measured in thecortical collecting tubule (CCT) of mice expressing truncated -subunitsof ENaC, reproducing one of the mutations found in human patients withLiddle's syndrome. Tubules were isolated from mice homozygous for the Liddlemutation (L/L) and from wild-type (WT) littermates. Amiloride-sensitivecurrents ( I Na ) from single cells were recorded under wholecell clamp conditions. CCTs from mice kept under control conditions and fed adiet with normal levels of Na had very small I Na s (WT: 18± 13 pA; L/L: 22 ± 8 pA at V m = -100 mV) thatwere not different in WT and L/L animals. However, the L/L mice had muchlarger currents when the animals were fed a low-Na diet (WT: 256 ± 127pA; L/L: 1,820 ± 330 pA) or infused with aldosterone (WT: 285 ±63 pA; L/L: 1,600 ± 280 pA). Currents from L/L mice were also largerwhen animals were pretreated with a high-K diet but not when the CCTs werestimulated in vitro with 8-CTP-cAMP. Noise analysis of amiloride-inducedfluctuations in I Na showed that single-channel currents atV m = 0 mV were slightly smaller in L/L mice (WT: 0.33 pA; L/L: 0.24pA). This difference could be attributed to a decrease in driving force sincecurrent-voltage analysis indicated that intracellular Na was increased in the L/L animals. Analysis of spontaneous channel noise indicated that the openprobability was similar in the two genotypes(WT: 0.77; L/L: 0.80). Thus theincrease in whole cell current is attributed to a difference in the density ofconducting channels.
      【关键词】 kidney sodium transport aldosterone hypertension amiloride
   ----------) I4 w  Q  ?- _8 b( r6 O6 Q

EPITHELIAL NA CHANNELS (ENaCs) are important in the regulation of renal salt excretion and hence to the regulation of blood pressure( 24 ). Liddle's syndrome, aMendelian form of hypertension, involves truncations or mutations of the COOHtermini of - or -subunits of ENaC( 13, 32 ). These alterationsincrease the current through ENaCs expressed in Xenopus laevis oo-cytes, a result due at least in part to an increase in surface expression of the subunits ( 6 ). Becausethe mutations usually affect a COOH-terminal PPPxY motif thought to act as aninternalization signal ( 13 ),it is possible that the increased expression is the result of a decreased rateof removal of channels from the plasma membrane. Recently, a mouse model ofLiddle's syndrome was developed( 24 ). When fed a high-Na diet,these animals have a phenotype that reproduces most of the features of the human syndrome.: ^- P/ e: |8 {
  R: e. P0 ?  z' x
In the cortical collecting tubule (CCT), Na channels are tightly regulatedby aldosterone. Tubules isolated from normal, salt-replete rats have almost nochannel activity as measured by transepithelial Na fluxes( 25, 31 ), single-channel( 20 ), or whole cell currents ( 7, 9 ). Channel activity isstrongly elevated by salt depletion or treatment with aldosterone. Thus lowlevels of mineralocorticoids can completely suppress the channels, at least inrodents. However, in humans with Liddle's syndrome, aldosterone levels inplasma are below normal. This suggests that in this disease mineralocorticoidcontrol of the channels is lost and channel activity is at least partlyaldosterone independent. The aim of the present study was to examine whetherthe channels are as strongly controlled in the mouse CCT and whether control was abolished or reduced in the animals with the Liddle's syndrome mutation.' t# p. M7 e6 n3 ?: A6 J

METHODS

Animals. Initial experiments were carried out using a commercial C57BL/6 mouse strain (Charles River Laboratories, Kingston, NY). Mice with the -ENaC subunit truncated at position R564 were generated as describedpreviously ( 24 ). For thepresent set of experiments, these animals were backcrossed onto the C57BL/6Jmouse background (N10). Mice homozygous mutant for the Liddle mutation (L/L)and their littermate controls were obtained by crossbreeding heterozygousmutant ( /L) mice. Genotyping was carried out by PCR as described( 24 ).

Animals were fed either a standard lab chow (Na content 2.8 g/kg; K content11 g/kg; Purina), a low-Na diet (Na content 3.8 mg/kg; K content 8.6 g/kg;ICN, Cleveland, OH), or a high-K diet (Na content 5 g/kg; KCl content 100g/kg; Harlan Teklad, Madison, WI). The low-Na diet was given for either 2 daysor for 6-8 days. The high-K diet was given for 6-8 days. Someanimals were infused with aldosterone via subcutaneous osmotic minipumps(Alzet model 1002, Alza, Palo Alto, CA) for 6-7 days. Aldosterone wasdissolved in polyethylene glycol-300 at a concentration of 2 mg/ml. Theinfusion rate was 10 µg/day.
0 }+ b' q  A2 W. ]8 m2 k
Tissue preparation. After the animals were killed, the kidneys were removed, and CCTs were dissected free and opened manually to expose theluminal surface. Under these conditions, the tissues retain their epithelialstructure and the cells are presumed to remain polarized. The split tubuleswere attached to a small plastic rectangle coated with Cell-Tak (Collaborative Research, Bedford, MA) and placed in a perfusion chamber mounted on aninverted microscope. The chamber was continuously perfused with solutionpre-warmed to 37°C consisting of (in mM) 135 Na methane sulfonate, 5 KCl,2 CaCl 2, 1 MgCl 2, 2 glucose, 5 BaCl 2, and 10HEPES adjusted to pH 7.4 with NaOH.

Electrophysiology. The patch-clamp pipettes were filled with solutions containing (in mM) 7 KCl, 123 aspartic acid, 20 CsOH, 20 TEAOH, 5EGTA, 10 HEPES, 3 MgATP, and 0.3 NaGDP S with the pH adjusted to 7.4 withKOH. Basic protocols for measuring whole cell amiloride-sensitive current werepreviously described ( 8, 9 ).
/ j3 F7 r, [& o0 T/ s% d$ E! V
For amiloride-induced noise analysis, the cell membrane potential wasclamped to 0 mV and currents were recorded at a gain of 10-100 pA/mVonto videotape using a digital data recorder (VR-10, Instrutech). The cellswere superfused with bath media containing no amiloride, one or moresubmaximal concentrations of amiloride (0.25-1.5 µM), and 10 µMamiloride, a dose considered to be maximum. Data were then filtered at 500 Hz,digitized at 1 kHz, and intervals of 10-30 s were transformed using thefast Fourier transform application within the PClamp 8 software package (AxonInstruments). The resulting spectra were fit with Lorentzians over a frequency range of 0.5 to 200 Hz

The single-channel current ( i ) was then computed from the fitparameters according to the relationship( 15 )
" d1 i7 F0 Q8 {1 N  P
where S o is the plateau of the Lorentzians, =2 f c where f c is the corner frequency, I Na is the macroscopic current through thechannels, and P A is the probability that the channel isblocked by amiloride and is given by
& h, x. Y& P0 P1 m
where I Na (A) and I Na (0) are theleak-corrected currents in the presence and absence of the submaximal dose ofamiloride.
2 R/ s7 L" p$ ~( _
To estimate the open probability ( P o ) of the channels,the digitized records in the absence of amiloride and in the presence of amaximal dose were converted to all-points histograms and fitted with Gaussianfunctions, also using PClamp 8 software. The variance of the currents throughthe channels ( Na ) was calculated from
: B& h! Z0 U) a  Q* H
where T 2 and A 2 are the variances of the currents in the absence and presence of 10 µMamiloride, estimated from the width of the fitted Gaussian functions. Thisrelationship assumes that fluctuations in currents through the channel( Na 2 ) and through leak pathways ( A 2 ) are independent. P o wasthen estimated from the relationship ( 10, 14 )( C0 C8 C* K8 o
4 ?! ~% U9 a' U
RESULTS8 G. k5 |( i% K* Z7 Q0 \
2 V0 \) ^: p7 K6 e% y: J
Previous patch-clamp measurements of Na channel activity in the collectingtubule were done in rat kidneys( 9, 22 ). We therefore carried outpilot experiments using the commercial mouse strain (C57BL/6) used to producethe mutant mice for the later studies to see whether similar results wereobtained. Figure 1 shows wholecell currents from a cell of a CCT isolated from a mouse that had been treatedwith aldosterone by an osmotic minipump for 7 days. Currents were obtained inthe absence and presence of 10 µM amiloride. Because the pipette contained no Na, there is a large Na gradient across the membrane. Under the conditionsused, amiloride reduced the inward current in the absence of a transmembranevoltage and blocked a large fraction of the total inward conductance. Theamiloride-sensitive currents approached zero at large positive potentials,consistent with blockade of Na-selective channels with low intracellular Na.At a cell potential of -100 mV relative to the bath, I Na was 246 ± 58 pA ( n = 16 cells). InCCTs from animals not treated with aldosterone, only 1 of 13 cells had a clear amiloride-sensitive current and the mean I Na was 13± 13 pA. These data are similar to those obtained in rat CCT ( 7, 9, 20 ). Na channel function inthese cells, at least as evaluated in vitro, was dependent on an elevation ofplasma mineralocorticoid activity.
! P0 U0 [9 {; f+ ]4 y# b
Fig. 1. Current-voltage ( I-V ) relationships in mouse cortical collectingtubule (CCT). Data were obtained from a wild-type C57BL/6 mouse infused withaldosterone for 7 days. A : currents were measured under whole cellvoltage-clamp conditions. The holding potential was 0, and the test potentialswere stepped from -120 to  60 mV in 10-mV increments. a : Tracesobtained in the absence of amiloride; b : traces set in the presenceof 10 µM amiloride. B : I-V curves were generated bymeasuring currents at the end of each voltage pulse in the absence ( )and presence ( ) of amiloride. The represent the difference currents( I Na )./ H0 p: `, I9 B( k" A( A2 k

We next examined the L/L mice and their wild-type (WT) littermates backcrossed into this same mouse strain. Because human patients with Liddle'ssyndrome have high blood pressure, presumably reflecting increased Nareabsorption, together with low levels of aldosterone, we hypothesized thatthe CCTs from the untreated L/L animals would escape from tightmineralocorticoid control and would therefore have high channel activity undercontrol conditions. Contrary to this expectation, I Na wasvery low in both the L/L and WT mice under these control conditions ( Fig. 2 ), and there was nostatistically significant difference between the two groups( Fig. 3 ).+ x/ y, F% ^, x9 K9 l0 {  p
  v! u( k  v; ~$ Y1 L2 j# s
Fig. 2. I-V relationships for wild-type (WT; normal; left ) andLiddle (L/L; right ) mouse CCT. Animals were fed either standard chow( bottom ) or a low-Na diet ( top ). In each case, I-V curves were obtained in the absence ( ) and presence ( ) of amilorideand I Na - V curves ( ) calculated from thedifference in current at each voltage.
. Y- x! U! F9 [  f$ [$ u% ?2 z
Fig. 3. Mean values of I Na measured at V m =-100 mV under different in vivo and in vitro conditions. Open barsrepresent WT mice. Filled bars represent Liddle mice. Control = standard chow,no infusion. Aldo = standard chow, aldosterone infusion by osmotic minipumpfor 6-8 days. Low Na = low-Na diet for 2 or 7 days (d). High K = high-Kdiet for 6-8 days. cAMP = CCTs from mice fed standard chow, withoutaldosterone infusion treated in vitro with 10 - 5 M8-cpt-cAMP for 5 to 30 min. Data are shown as means ± SE for 11 to 23different cells. Statistical tests for Liddle vs. WT littermates: * P + J3 P0 t0 k  v( u: ?+ B: ^
# v' Q; e3 ?0 M' ^
We next tested whether a difference could be detected after activation ofchannels with aldosterone. Figure2 shows representative current-voltage ( I-V ) curves fromaldosterone-infused mice. The WT animals had Na currents that were similar tothose from the standard mouse strain described above. In contrast, the currents from tubules taken from L/L mice were considerably larger. In thecell shown, currents could not even be measured at voltages more negative than-60 mV due to saturation of the amplifier. In addition, the reversalpotential of the amiloride-sensitive current was shifted toward zero. Thepipette solution contained no Na, so the most likely interpretation is that Naentered the cells through the channels so rapidly that the submembrane spacecontained a significant Na concentration leading to a finite reversalpotential. Mean values for I Na under these conditions aregiven in Fig. 3. Values for the L/L cells are more than five times larger than those from the WT cells, andthe differences are highly significant. Thus, although the basal currents inthe two groups were not different, the L/L animals responded much morevigorously to elevation of plasma aldosterone than did controls.( F* n" ^: z8 `) b- Y

Reduction in dietary Na intake is a more physiological way of elevatingmineralocorticoid status. WT mice that were fed a low-Na diet for 6-8days also had easily measurable I Na s that were comparableto those obtained with aldosterone infusion ( Fig. 3 ). Again, the L/Lanimals had much larger currents than did the WT animals, the average valuesbeing increased by a factor of 7. We also investigated animals after 2 days ofNa depletion. This is sufficient to activate I Na but notto the presumably maximal levels seen after 1 wk of the low-Na diet. Theresponse of the L/L animals was again significantly larger than that of the WT( Fig. 3 ).
0 n8 E4 l$ ]- S
To see whether channels regulated through other physiological pathways werealso hyperactivated in the L/L mice, we fed mice a high-K diet for 6-8days. Although plasma aldosterone can be elevated under these conditions, weobserved enhanced Na channel activity even in adrenalectomized rats,suggesting a process that is at least in part aldosterone independent ( 21 ). The high-K dietincreased I Na in both WT and L/L mice, although to alesser extent than did the low-Na diet or aldosterone infusion. The meancurrents in the L/L animals were significantly larger, although the ratio ofL/L to WT was only 2.3, less than that for the other maneuvers( Fig. 3 ).2 X/ a: q7 U, m5 w) p2 \

Na channels in the CCT can be activated in vitro by raising cytoplasmiccAMP ( 8 ). We compared theresponse to an application of 8-CTP-cAMP in CCTs from WT and L/L mice. Asshown in Fig. 3, 8-CTP-cAMPincreased I Na in both sets of animals, and there was nodifference in the mean levels of the two groups. We conclude that the effectsof the Liddle mutation on Na channels depend on the pathway (i.e.,mineralocorticoids vs. cAMP) through which the channels are activated., l3 ?) g' |6 S
5 d; Q# d) w6 p. R, d
To examine the single-channel properties of the WT and L/L channels, wefirst examined cell-attached patches. The results were disappointing. Sealsgood enough for single-channel analysis were difficult to make regularly, andin those patches where the seal was satisfactory, the channels were largelyabsent. In many cases, channels were seen briefly after seal formation butinactivated quickly. Thus the mouse CCT is not as advantageous as that of therat for investigating single-channel events.7 g* o" o' c  A
, I& }: o9 N- x$ W) ], B
Because good whole cell recordings could be made relatively routinely, wechose to further characterize the channels using noise analysis. Two types ofexperiments were done. First, we used analysis of amiloride-inducedfluctuations ( 14, 15 ) to estimate thesingle-channel current. We then analyzed noise from spontaneous fluctuationsin channel currents to estimate the P o of thechannels.: I3 D& N9 ]- [+ P
& X* n  T0 w( H+ u2 P! B5 F2 g+ H7 @
Spectra from a representative whole cell recording in the absence ofamiloride and in the presence of submaximal (1 µM) and maximal (10 µM)amiloride concentrations are illustrated in Fig. 4. In the absence ofamiloride, significant power densities were seen over a large range offrequencies. The shape of the power density spectrum is complex, however, and was not analyzed further. With 1 µM amiloride, a clear Lorentzian componentof noise emerged with a plateau in the frequency range of 0.1 to 1 Hz. Theentire spectrum from 0.1 to 100 Hz could be well described by a singleLorentzian component with a plateau of 3.06 pA 2 s and a cornerfrequency of 3.8 Hz. Finally, in the presence of 10 µM amiloride, the noise in the range of 0.1 to 10 Hz was greatly reduced. The spectrum between 1 and300 Hz could also be reasonably well described by a Lorentzian with a muchreduced plateau of 0.06 pA 2 s and an increased corner frequency of42 Hz.
7 o; |3 L* m$ v4 r( B$ r
Fig. 4. Power density spectra of currents in mouse CCT. The examples shown are froma L/L mouse infused with aldosterone. The cell membrane was clamped to 0 mVand currents were recorded for 10-20 s with 0 ( A ), 1( B ), and 10 µM amiloride ( C ) in the bath. Power densityspectra were obtained by fast Fourier transform. The line through the spectrumwith 1 µM amiloride is the best fit to a Lorentzian function withS o = 3.06 pA 2 s and f c = 3.8 Hz.

The corner frequency of the amiloride-induced noise is expected to increasewith amiloride concentration. This is shown in Fig. 5, which shows data fromboth WT and L/L animals. The results from the two genotypes are identical,implying that the kinetics of amiloride block are unchanged by the mutation. The slope of the plot gives the on-rate for block, which was 56 µM/s. Theintercept with the frequency axis gives the off-rate, but this was too closeto zero to be accurately determined.
; n5 d2 }4 A) ?# z  ~4 V( B- @- n
Fig. 5. Kinetics of amiloride block. A : dose-response curves foramiloride-sensitive currents. I Na measured in the presenceof submaximal amiloride concentrations is plotted as a fraction of the currentin the absence of amiloride for Liddle ( ) and normal ( ) mice. Allanimals were infused with aldosterone. The line represents the best fit of thedata with Liddle's mice to a simple blocking scheme with K i = 0.25 µM. B : corner frequency as afunction of amiloride concentration. Data were obtained by fitting powerdensity spectra with Lorentzian functions for Liddle ( ) or normal( ) mice. The line is a best fit to the equation with k on = 56 µM/s and k off =5.6/s.  ^  l: H8 a# D$ w

These data can be used to estimate single-channel currents from Eq.1. For each cell studied, we calculated i (at a membrane potential of 0) for one to three submaximal amiloride concentrations. Althoughthere was some variation from cell to cell, values obtained using differentconcentrations on the same cell were usually in good agreement. The meanvalues for i were 0.33 pA for WT and 0.24 pA for L/L( Table 1 ). The reduced values for the mutant channels can be at least partially attributed to higherintracellular Na concentrations as shown in Fig. 2 (see DISCUSSION ). In any case, the higher whole cell currents in the L/Lcells are not due to an increased single-channel current.7 ?2 ]3 K) c! s% g; h6 q  _: J

Table 1. Effect of Liddle mutation on single-channel parameters estimated fromfluctuation analysis

Channel P o can be estimated assuming that the noise inthe current traces in the absence of amiloride arises from spontaneous transitions of channels between open and closed states. As shown in Fig. 6, distributions ofcurrents about the mean could be described by Gaussian functions both in theabsence of amiloride and in the presence of a maximal amiloride dose. Thispermits the variance of the currents to be estimated from the width of theGaussian fit. Assuming that the noise measured in the presence of amiloriderepresents current fluctuations unrelated to Na channels, the variance of I Na can be calculated from the relationship Na 2 = T 2 - A 2 as described in the METHODS section.
7 G' t- W& T2 W+ b
Fig. 6. Spontaneous current fluctuations. Current histograms were obtained from thesame cell used for Fig. 4 inthe absence ( A ) and presence ( B )of10 µM amiloride.Histograms were fitted with Gaussian functions Without amiloride Î = 470 pA, = 3.73 pA. Withamiloride, Î = 122 pA, = 2.45 pA.& I/ B6 e9 n- i9 k9 `

Equation 2 shows the relationship between the variance of the current around the mean and the single-channel parameters. The mean values for P o calculated in this way were quite high: 0.77 for WT and0.80 for L/L ( Table 1 ). Thusthe higher currents in the L/L cells appear to be the result of an increase in the number of conducting channels ( N ) rather than to a change in P o or i. Limitations to this interpretation arediscussed below.  M" @2 a1 j, l& R

DISCUSSION/ N. n7 ?1 f: }: ^

Na channels in mouse CCT. The results shown in Figs. 1 and 2 indicate that in the mouseCCT control of Na channels by aldosterone is very tight. In most cases, noclear amiloride-sensitive current could be observed in control animals fed astandard rodent chow. Typically, the sensitivity of the assay was such that acurrent of 10-20 pA at -100 mV would have been detectable. Incontrast, most cells had measurable Na currents after elevation of plasmaaldosterone either directly through infusion or indirectly by decreasingdietary Na intake. In this sense, the mouse is similar to the rat( 7, 20, 25, 31 ) but is unlike the rabbit.In the latter case, there appears to be a constitutive channel activity thatgives rise to an amiloride-sensitive Na conductance and active Na transport even when the mineralocorticoid status is low( 19, 27 ).
! V  K) C9 a4 z! [( `8 a  [% M
Augmented channel activity in CCTs from Liddle mice. Under conditions in which Na channels are activated in the CCT in vivo, particularlyby Na depletion or aldosterone infusion, the amiloride-sensitive currents werefive to seven times higher in the L/L animals compared with their WT controls.This degree of stimulation is similar to that observed in the oocyte system, in which injection of comparable amounts of cRNA for the WT and truncated -subunit along with normal - and -ENaC generate amiloride-sensitive currents that are generally two to five times larger inthe case of the mutant channels( 1, 6, 26, 29 ).. o+ V* n( E/ D+ Q, K  ]
+ q5 d0 D% q- t
The channel activity in the L/L cells is quantitatively rather thanqualitatively altered. The single-channel currents at a membrane potential ofzero were slightly smaller for the L/L channels, but this can be explained atleast in part by a smaller Na gradient across the membrane. It appears( Fig. 2 ) that when the inwardNa currents are very high, as in the case of the L/L cells, that theintracellular Na concentration at the inside of the cell membrane can befairly high even though the pipette solution in communication with thecytoplasm is Na free. The typical reversal potential of theamiloride-sensitive I-V relationship was about -35 mV. Assumingthat the channels remain highly selective for Na, this corresponds to asubmembrane Na concentration of 38 mM. Given the very large fluxes acrossthe apical membrane under these conditions, this value seems plausible,although we cannot completely rule out the possibility that the change inreversal potential could reflect a reduction in Na/K or Na/Cs selectivity ofthe channel. Thus the gradient for Na across the membrane could be reduced from 140 to 100 mM, which could account for the lower single-channel currents. This finding is in agreement with single-channel recordings ofchannels with Liddle mutations in oocytes( 6, 26, 29 ).& j( e3 [* [! w7 @4 U# q. [0 ]

The open probability, again estimated from noise analysis, was also similarfor WT and L/L channels. This analysis assumes a single population of channelswith the same P o that may not be the case( 23 ). Because channels withvery low P o would contribute little to either the noise orthe current, conversion of such channels to a high P o formwould be indistinguishable from a simple increase in channel number at aconstant P o. However, the results rule out a uniformincrease in P o as a major mechanism for the effects of themutation.
0 z# Q8 q3 T' f4 e
In the oocyte system, Firsov et al.( 6 ) found that the increased I Na could be only partially accounted for by an increasein the amount of protein expressed at the cell surface. P o, measured as the ratio of the macroscopic current tothe product of the channel protein and the single-channel currents, was quite low (8 ?- W0 J4 a+ o* p

The inhibition constant for the interaction of the channels with amilorideestimated from dose-response curves was 0.25 µM at V m = 0( Fig. 5 A ). The on-ratefor amiloride block, calculated from the slope of the plot of corner frequencyvs. concentration, was 56 µM/s. In principle, the off-rate can bedetermined as the intercept of this plot with the frequency axis. The best fitgives an estimate of 5.6/s, but this is not very reliable because small errorsin the estimate of the slope would have large effects on the intercept that isclose to the origin. A better estimate of the off-rate can be obtained bydividing the on-rate by the inhibition constant, giving a value of 14/s. Therewas no difference in the kinetics of amiloride block of WT and L/Lchannels.

Unfortunately, we could not routinely record single-channel events from theapical membrane of the mouse CCT. Several problems appeared to contribute tothis failure. First, many seals, although sufficiently tight to achieve goodwhole cell clamps, had resistances that were too low to resolve single-channel events. Second, patches for which the seal was adequate were often unstablewhen a voltage was applied to the pipette. Third, in a number of patches inwhich activity was visible immediately after seal formation, the channelsinactivated spontaneously before a recording could be made. Finally, patcheswith high resistance and stable seals tended to have no channel activity atall. Although we could occasionally obtain data similar to those seen in therat CCT, we could not be sure how representative such recordings were andtherefore have not reported them here.

Requirement for activation. The most surprising finding to us wasthe requirement for conditions in which channels are normally activated, e.g.,high plasma aldosterone, to observe the increased currents associated with theLiddle mutation. Indeed, our hypothesis when we began the work was that oneconsequence of the mutation would be a loss of mineralocorticoid control ofthe channels. We therefore expected that currents under control conditions would be larger in the L/L animals than in WT. Furthermore, if the main effectof the mutation was to bypass this control mechanism, it is possible that thecurrents under conditions of maximal stimulation might not be very different.In this sense, the channel activity would reflect the state ofpseudohyperaldosteronism, which characterizes patients with Liddle's syndrome( 13, 32 ). The observations wereopposite to these expectations. Under control conditions, channel activityremained suppressed, and they increased to a much greater extent in responseto stimulation. They also differ from results obtained from primary cultures derived from CCTs isolated from WT and L/L mice, in which a significantamiloride-sensitive short-circuit current could be measured in the absence ofaldosterone, and this current was larger in L/L than in WT mice (Pradervand S,Vandewalle A, Bens M, Gautschi I, Loffing J, Hummler E, Schild L, and Rossier BC, unpublished observations). Thus the ability of the cell to suppresschannel activity when aldosterone levels are low appears to be stronger invivo than in vitro.
9 p  T  g% f% Q5 o# X% ~  Y
Not every mode of stimulation resulted in larger currents in the L/L mice.The largest differences from WT were observed with high plasma aldosterone,achieved either with a low-Na diet or with direct infusion of the hormone.Significantly higher currents were also observed when the animals were given ahigh-K diet. Although increased K intake can also stimulate aldosteronesecretion ( 3 ), there is alsoevidence that at least some of the effects of high K are independent ofaldosterone ( 18, 30, 33 ). In particular, thestimulation of Na channels in the CCTs of rats on a high-K diet was unaffectedby adrenalectomy ( 21 ). Thusthe Na currents measured under these conditions may be only partly induced byaldosterone. This could explain the smaller degree of hyperactivation in theL/L animals, but experiments with adrenalectomized WT and L/L mice would be necessary to test this point rigorously.  Y1 \! }3 C; d) q: Z

In contrast, stimulation of I Na in vitro using cAMPevoked similar currents in WT and L/L genotypes. Increased intracellular cAMPis believed to underlie the activation of channels by ADH and perhaps by otherhormones using this second messenger system( 12 ). With the use of anepithelial exogenous expression system, Snyder( 28 ) found that cAMP-activatedcurrents were actually lower in Liddle mutant channels than with WT channels. Although we did not find an inhibition in the cAMP response in the nativekidney cells, we did confirm a striking difference in the effects of theLiddle mutation on cAMP-stimulated channels compared with those activatedthrough other pathways.+ y+ h" w; j7 X( s, l1 t

The increase in activity of Liddle channels in oocytes was dependent on therate of Na influx across the plasma membrane, suggesting that the mutantchannels had escaped from downregulation due to high intracellular Naconcentration ( 11 ). Such amechanism may also be involved in the experiments reported here in the CCT.Such an effect could not involve an acute response to high cell Na becauseunder the conditions of measurement, the concentrations are very low, at leastin the control cells. A more chronic effect is more likely. However, similarresults were obtained in animals on a low-Na diet, where Na delivery to theCCT (and Na influx into the cells) is low, and in animals infused withaldosterone in which Na delivery is presumably high. Thus the effects of themutation do not appear to depend entirely on an elevation of intracellular Nain vivo. Nevertheless, a different response to increased cell Na couldcontribute to the findings.

Implications for hypertension. Liddle's syndrome is a form of pseudohyperaldosteronism. Patients have symptoms of primary aldosteronism,including hypertension, hypokalemia, and metabolic alkalosis, but actualplasma levels of the hormone are low ( 32 ). If low-aldosteronelevels can suppress the activity of Na channels even with a Liddle mutation,as appears to be the case at least in the mouse CCT, how can high bloodpressure develop?

One possibility is that channels are suppressed more effectively in therodent than in the human. As mentioned above, the rabbit CCT has a significantconstitutive (aldosterone independent) rate of Na transport. If this were alsothe case in the human kidney, the constitutive channels might also behyperactive in Liddle's syndrome. According to this explanation, hypertension should be less pronounced in the L/L mouse than in the human disease. Thisappears to be the case as blood pressure in these animals was only mildlyelevated and only when dietary Na was high( 24 ).

A second possibility is that Na channels in other nephron segments are lesscompletely suppressed by low aldosterone in both mice and men. One candidatefor such a segment is the connecting tubule (CNT), where at least in therabbit Na transport rates were higher than in the CCT and less sensitive tomineralocorticoid status ( 2 ).In this scenario, hyperreabsorption of Na would take place in the CNT ratherthan the CCT. No data are available for mouse or rat CNT in vitro. However, asubstantial amiloride-sensitive transport rate was observed in the late distalconvoluted tubule of control rats by micropuncture techniques( 4 ).0 e+ i" J4 {0 x; W. i& ^

Third, it is possible that channels in any of the distal segments could beactive in vivo due to the presence of hormones that are eliminated when thetubules are studied in vitro. The effects of such hormones could beexaggerated in Liddle's syndrome, leading to overreabsorption of Na. Anobvious candidate is ADH, whose action is rapidly reversed and therefore not"remembered" by the in vitro tissue. However, activation by cAMP,the second messenger for ADH, was not elevated in the L/L mouse CCT. We cannot, however, rule out a role for other hormone-second messenger systems.2 x! z* r; }2 N0 q( d3 h

Implications for aldosterone action. A clear result of these studies is that the channels with the Liddle mutation can be regulated byaldosterone. This finding indicates that an intact COOH terminus of the -ENaC subunit, and its PPPxY motif presumed to be an internalizationsignal, is not required for the long-term effects of mineralocorticoids.Recently, Staub and colleagues ( 1 ) proposed a mechanism ofaction of aldosterone involving a phosphorylation of the COOH-terminal bindingprotein Nedd-4-2 by serum and glucocorticoid activated kinase (SGK) witha subsequent impairment of channel ubiquitination and retrieval from the apical membrane. This model predicts that mineralocorticoid control of thechannels should be diminished in the absence of the internalization signal,which we did not observe. Two caveats should be mentioned. First, Abriel etal. ( 1 ) used channels in whichthe PPPxY motif was eliminated from both the - and -subunits,whereas in the Liddle mice only the -subunit was altered. Second, theSGK-mediated pathway may be important in the rapid (1-3 h) effects ofthe hormone, whereas 2 days) exposure tomineralocorticoids and/or salt depletion.# X# B! U! a8 R/ B) Q" e6 k! v8 ?
, a5 Y4 B( X; z% S5 g
Because the long-term effects of aldosterone appear to involve aredistribution of channel protein from cytoplasmic to plasma membrane sites( 16, 17 ), one possibility is thataldosterone could stimulate the rate of insertion of channels into the apical membrane. This would increase the steady-state density of channels availablefor transport, and the effect would be exaggerated if the rate of channelretrieval were at the same time diminished, as may occur in Liddle'ssyndrome.

DISCLOSURES

This work was supported by National Institute of Diabetes and Digestive andKidney Diseases Grant DK-59659, a grant-in-aid from the American HeartAssociation Heritage Affiliate, and a Fellowship Grant Da536/1-1 fromthe Deutsche Forschungsgemeinschaft (to A. Dahlmann).
      【参考文献】! Y( B3 ^5 {( {* L8 m
Abriel H,Loffing J, Rebhun JF, Pratt JH, Schild L, Horisberger JD, Rotin D, and StaubO. Defective regulation of the epithelial Na channel by Nedd4in Liddle's Syndrome. J Clin Invest 103: 667-673,1999." a' A2 J; O; ~2 f) {4 d
8 P  Z( i' b! T* X) p
5 ?) o4 T/ E5 d+ [) F
' a# f" Z" z* S# c) G7 U. z
Almeida AJ andBurg MB. Sodium transport in the rabbit connecting tubule. Am JPhysiol Renal Fluid Electrolyte Physiol 243: F330-F334,1982.# J7 T( g( h  I' H4 d4 ]
- G! G( _2 H- ^1 b  ]& U8 D" U

Boyd JE,Palmore WP, and Mulrow PJ. Role of potassium in the control of aldosteronesecretion in the rat. Endocrinology 88: 556-565,1971.

5 q% }% J8 |( y% h8 k. l1 q8 x: Q
Costanzo LS. Comparison of calcium and sodium transport inearly and late rat distal tubules: effect of amiloride. Am JPhysiol Renal Fluid Electrolyte Physiol 246: F937-F945,1984.6 `5 |7 |! w$ F" L

Debonneville C,Flores SY, Kamynina E, Plant PJ, Tauxe C, Thomas MA, Münster C, ChraibiA, Pratt JH, Horisberger JD, Pearce D, Loffing J, and Staub O. Phosphorylation of Nedd4-2 by Sgk1 regulates epithelial Na channel cell surface expression. EMBO J 20: 7052-7059,2001.
4 G0 E; ^4 o) ~3 F

Firsov D,Schild L, Gautschi I, Merillat AM, Schneeberger E, and Rossier BC. Cellsurface expression of the epithelial Na channel and a mutant causing Liddlesyndrome: a quantitative approach. Proc Natl Acad SciUSA 93:15370-15375, 1996.
. i* o; v% [; l* u% L
, b% V" G; t! G- i
9 @& C4 b1 M. O8 S9 [1 T4 y- S7 [
Frindt G,Masilamani S, Knepper MA, and Palmer LG. Activation of epithelial Nachannels during short-term Na deprivation. Am J Physiol RenalPhysiol 280:F112-F118, 2001.8 e0 |2 w' i( z" [4 i5 D# H
! ]9 l* q9 n1 K# [: f4 ]6 p& q
  M; O1 h: Q: F" o1 e
8 L' ?2 w8 j" a. `$ c3 ^' V
Frindt G andPalmer LG. Regulation of Na channels in the rat cortical collectingtubule: effects of cAMP and methyl donors. Am J Physiol Renal FluidElectrolyte Physiol 271:F1086-F1092, 1996.; T0 d/ R/ `. f5 r# i8 S
$ y# Q- C3 f& e7 w* G1 X# t
: K& o5 ?! D! y3 H
% a4 j, I+ a1 X
Frindt G,Sackin H, and Palmer LG. Whole-cell currents in rat cortical collectingtubule: low-Na diet increases amiloride-sensitive conductance. Am JPhysiol Renal Fluid Electrolyte Physiol 258: F562-F567,1990.6 T$ x3 e9 b& D' O8 {
, t4 I1 V: t5 B$ h2 ?4 B+ I; E' X# t

Frindt G,Silver RB, Windhager EE, and Palmer LG. Feedback regulation of Na channelsin rat CCT. III. Response to cAMP. Am J Physiol Renal FluidElectrolyte Physiol 268:F480-F489, 1995.. s. w1 k  l) k0 V

+ j2 |5 ^2 b) q, g" h! C6 }, @
% Q# c- r3 [8 N  z: k. Q
Kellenberger S,Gautschi I, Rossier BC, and Schild L. Mutations causing Liddle syndromereduce sodium-dependent downregulation of the epithelial sodium channel in the Xenopus oocyte expression system. J ClinInvest 101:2741-2750, 1998.

Kurokawa K,Fukagawa M, Hayashi M, and Saruta T. Renal receptors and cellularmechanisms of hormone action in the kidney. In: The Kidney

hysiology and Pathophysiology, edited by Seldin DW and GiebischG. New York: Raven, 1992, p.1339-1372.

# A  _2 b/ M- o8 Q* I. k& t/ Q; p: m
Lifton RP,Gharavi AG, and Geller DS. Molecular mechanisms of human hypertension. Cell 104:545-556, 2001.7 Y3 o- u0 t( ]9 `, `. t4 A! X3 ^+ ?: u
4 J) R. L9 B& s9 b- F! V, Y% x3 g

& H' B( N8 [- U& b& _
Lindemann B. The beginning of fluctuation analysis ofepithelial ion transport. J Membr Biol 54: 1-11,1980.: B. W( \# L/ S8 }* _9 p

# F1 s; N/ W8 |7 `0 ?, k
Lindemann B andVan-Driessche W. Sodium specific membrane channels in frog skin are pores:current fluctuations reveal high turnover. Science 195: 292-294,1977.
: x8 M3 R; Y0 Z; K' e* e
' v  N& ]' {  k5 v1 i/ t

Loffing J,Pietri L, Aregger F, Bloch-Faure M, Ziegler U, Meneton P, Rossier BC, andKaissling B. Differential sub-cellular localization of ENaC subunits inmouse kidney in response to high- and low-Na diets. Am J PhysiolRenal Physiol 279:F252-F258, 2000.. o: o9 C8 L  Z+ F; o, Q( p
- e4 ~; z- p- Y" I! f
( [0 Z0 O# ^3 e& J, K8 w2 S
2 Z/ k; E5 q/ K
Masilamani S,Kim GH, Mitchell C, Wade JB, and Knepper MA. Aldosterone-mediatedregulation of ENaC,, and subunit proteins in ratkidney. J Clin Invest 104:R19-R23, 1999.$ b4 W; s3 j6 S* q

% ?" X7 S% [& J6 S  _+ r
Muto S, SansomS, and Giebisch G. Effects of a high potassium diet on electricalproperties of cortical collecting ducts from adrenalectomized rabbits. J Clin Invest 81:376-380, 1988.. h, Z; l8 z5 x) ?2 ^1 B9 A
/ m( Z; D* H; e# N
: Z+ y, n/ Z6 c$ [$ [
/ s- u. m9 |3 g7 h& w
O'Neil RG. Aldosterone regulation of sodium and potassium transport in the corticalcollecting duct. Semin Nephrol 10: 365-374,1990.9 T- Y" ^4 z4 U! I
; N+ F0 ]- V( B0 E6 `# S6 }' `

1 C4 k+ P; m& K- z# t# l
Pácha J,Frindt G, Antonian L, Silver R, and Palmer LG. Regulation of Na channelsof the rat cortical collecting tubule by aldosterone. J GenPhysiol 102:25-42, 1993.
8 _% ?! _1 B: I# o& p, r) D2 h

; b: k6 x# a8 f$ x
Palmer LG,Antonian L, and Frindt G. Regulation of apical K and Na channels and Na/Kpumps in rat cortical collecting tubule by dietary K. J GenPhysiol 104:693-710, 1994.( O, j. L! N; c8 G" B- b
7 Y1 B+ e& \2 V3 O: r) D

Palmer LG andFrindt G. Amiloride-sensitive Na channels from the apical membrane of therat cortical collecting tubule. Proc Natl Acad Sci USA 83: 2767-2770,1986.7 W% u0 D+ C, `; U0 c
- x8 ?- O) Y5 E% u0 B( a0 }

% q7 [3 l- `) |1 N
Palmer LG andFrindt G. Gating of Na channels in the rat cortical collecting tubule:effects of voltage and membrane stretch. J Gen Physiol 107: 35-45,1996.2 M! P' ~) v% Z3 E0 L+ l
8 W6 H8 A$ I' Z4 W

Pradervand S,Wang Q, Burnier M, Beermann F, Horisberger JD, Hummler E, and Rossier BC. A mouse model for Liddle's Syndrome. J Am Soc Nephrol 10: 2527-2533,1999.
7 U2 Z4 b" K( B* U
% x1 d  M) P# ~: w- P4 w8 e
) b/ m9 q. w1 F* B1 S
Reif MC,Troutman SL, and Schafer JA. Sodium transport by rat cortical collectingtubule. Effects of vasopressin and desoxycorticosterone. J ClinInvest 77:1291-1298, 1986.
- [7 S; f6 k( J# u. E" G8 `2 C
4 }& s' @3 B/ y; O) J/ g: Z
) B. J) n- G5 D1 a
Schild L,Canessa CM, Shimkets DG, Warnock DG, Lifton RP, and Rossier BC. A mutationin the epithelial sodium channel causing Liddle's disease increases channelactivity in the Xenopus oocyte expression system. ProcNatl Acad Sci USA 92:5699-5703, 1995.
- }; ]  J; R" I6 W

Schwartz GJ andBurg MB. Mineralocorticoid effects on cation transport by corticalcollecting tubules in vitro. Am J Physiol Renal Fluid ElectrolytePhysiol 235:F576-F585, 1978.# G7 \! W& f; _; L0 p

Snyder PM. Liddle's syndrome mutations disrupt cAMP-mediated translocation of theepithelial Na channel to the cell surface. J ClinInvest 105:45-53, 2000.
$ b; `0 o) Z& w+ t" U; N

Snyder PM,Price MP, Mcdonald FJ, Adams CM, Volk KA, Zeiher BG, Stokes JB, and WelshMJ. Mechanism by which Liddle's syndrome mutations increase activity of ahuman epithelial Na channel. Cell 83: 969-978,1995.

' b$ K0 h/ U# R8 R6 t
) ]# ?8 p; _2 D5 E. Z, M) E
Stanton B, PanL, Deetjen H, Guckian V, and Giebisch G. Independent effects ofaldosterone and potassium on induction of potassium adaptation in rat kidney. J Clin Invest 79:198-206, 1987./ k+ L' M0 B5 V: x7 A& R& X

7 C- O8 @% _' F. ~
Tomita K,Pisano JJ, and Knepper MA. Control of sodium and potassium transport inthe cortical collecting tubule of the rat. Effects of bradykinin, vasopressin,and deoxycorticosterone. J Clin Invest 76: 132-136,1985.6 e' x3 [4 A5 l# ]" D

7 q& T) M  c* k$ D
8 \, }( v  m) l
Warnock DG. Liddle Syndrome: an autosomal dominant form of human hypertension. Kidney Int 53:18-24, 1998.2 h+ }) O. y# P( U: L6 A7 y. M
) P5 _1 Q( O7 u3 G7 n* m

Wingo CS,Seldin DW, Kokko J, and Jacobsen HR. Dietary modulation of activepotassium secretion in the cortical collecting tubule of adrenalectomizedrabbits. J Clin Invest 70:579-586, 1982.

收藏0 分享0 顶0 踩0