Electrophysiology of Triturus nephron: Cable properties and electrogenic transport systems

Kidney International, Vol. 37 (1990), PP. 157—1 70

ROBERT F. PITTS MEMORIAL LECTURE

Electrophysiology of Triturus nephron: Cable properties and electrogenic transport systems TAKESHI HosHI Laboratory of Physiology, School of Food and Nutritional Sciences, University of Shizuoka, Shizuoka 422, Japan The Robert F. Pitts Memorial Lectureships were founded in 1978, when the many friends and admirers of Robert Pitts established a fund in memory of that distinguished physiologist. The Lecturer is selected by the Renal Commission of the International Union of Physiological Sciences, with the Chair of Physiology at Cornell University Medical College acting ex officio. The Lecture is presented at each International Congress of Physiological Sciences. At the XXXI Congress held in Helsinki, Finland, the Pitts Lecturer was Professor Takeshi Hoshi, from Shizuoka, Japan. He delivered the lecture on July 13, 1989. There are several aspects about the choice of Proftssor Hoshi that seem singularly fitting. First and foremost, he is an outstanding physiologist who, like Robert Pitts, has contributed much to our understanding of renal function. Professor Hoshi, in a classical experiment with Sakai in the early sixties, showed for the first time that ions can cross the proximal tubule epit helium while avoiding the cytosol. These experiments which were concomitant with those of Windhager, Boulpaep and Giebisch, have been confirmed by others. The paracellular pathway for ion flow is now accepted as of paramount importance to the function of leaky epithelia. Then in 1968 Professor Hoshi with his coworkers discovered that glucose and sodium ions share a common mechanism as they enter the proximal tubular cells from the lumen during absorption. And shortly after he made similar findings concerning the relation between amino acids and sodium transport, and concerning the transport of other organic molecules and sodium. Again, these findings have been confirmed in other laboratories. These and other findings of Professor Hoshi contribute not only to a better knowledge of kidney function, but they have increased our understanding of the function of other epithelia as well. Professor Hoshi graduated from the School of Medicine of Tohoku University, Senday, and finished the Graduate Course in Physiology in 1954 in the same University, where he was an Instructor in Physiology until 1957 when he moved to the University of Tokyo. There in 1968 he became Professor of Physiology and Chairman until 1987. He then moved as Dean and Professor Emeritus of Physiology in the University of Shizuoka, Shizuoka. Professor Hoshi has been a Fuibright Scholar in the Laboratory of Professor Bodil Schmidt-Nielsen in Duke University in the United States. He has been scientific adviser to the Ministry of Education, Science and Culture of Japan. He is a member of the editorial board of various Journals and has held offices in the directive bodies of several Japanese scientific societies. Guillermo Whittembury Chair, Renal Commission IUPS

Previous Robert F. Pitts Memorial lectures 1980 RoIf Kinne, Federal Republic of Germany 1983 James A. Schafer, USA 1986 Gerhard Malnic, Brazil

As stated by R.F. Pitts [1] in his textbook, Physiology of the soon, the same methodology proved useful in the study of Kidney and Body Fluids, the kidney is a mysterious organ, not mechanisms of electrogenic nonelectrolyte transport. Initially, only with regards to its physiologic function and how it accom- two groups of investigators started such electrophysiological plishes this role, but also why it functions in this manner. studies independently: one at Cornell University [2, 3] and the Clearance methods and micropuncture techniques have played other at the University of Tokyo [4, 5]. These groups used major roles in the development of our appreciation of kidney amphibian (Necturus and Triturus) kidneys in their studies. function. In discovering how the kidney accomplishes its func- Later, this methodology was also applied to mammalian nephtion, particularly with respect to mechanisms of tubular func- rons [6]. At the present time, electrical measurements have tion, the introduction of new microtechniques has been useful become a routine technique in the study of nephron functions in in obtaining quantitative data from cells and cell membranes. many laboratories. Besides classical microelectrode techElectrophysiological methods using intracellular microelec- niques, more advanced techniques for measuring intracellular trodes were introduced to renal physiology about 30 years ago ionic activities and channel currents have contributed to our in an attempt to explore conductance properties of tubular cell understanding of nephron electrophysiology. Excellent reviews membranes and the tubular wall, as well as to measure the on special problems in the application of electrophysiological driving force behind ionic movements. Early studies with this methods to the study of nephrons have appeared previously methodology were mainly concerned with ion transport, but [7—10]. In this paper, I will discuss some aspects of electrical Received for publication and accepted July 31, 1989

properties of the nephron studied by cable analysis and the mechanisms of electrogenic (rheogenic) transport of organic solutes. This discussion is based on findings obtained by our

© 1990 by the International Society of Nephrology

group in the Triturus kidney. 157

Hoshi: Pills Memorial Lecture

158

Ta 130

220

330

400 tIm

Fig. 1. An example of intracellular voltage attenuation along the proximal tubule of Triturus kidney. Current pulses of a constant strength (6 x io- A) and 50 msec duration were passed intracellularly, and resultant electrotonic potentials were recorded at various distances (shown at the bottom of each set of tracings) from the injection site. Two sweeps were superposed for the outward-going (depolarizing) and the inward-going (hyperpolarizing) current. The upper beam indicates the duration and strength of the current pulse passed (only the depo-

larizing current was displayed), and the time signals of 10 msec intervals. Modified from [151.

the cell membrane. The calculated values of the specific resistances of the surface cell membrane (Rm) and the cytoplasmic core (R1) were 836 0cm2 and 625 0cm, respectively [15]. These estimates were thought to contain some error due to a structural simplification of the tubule and a lack of appreciation of the radial current flow in the vicinity of the current source. Nevertheless, the Rm value seemed sufficiently high to exclude the notion that the surface cell membrane was unusually leaky to ions. Since cellular interdigitation and the brush border structure were not taken into consideration in these calculations, the value was not thought to differ greatly from those of invertebrate nerve fibers [17] and mammalian cardiac purkinje fibers [18].

The calculated value of the cytoplasmic core resistance was two or three times higher than the specific resistance of the cytoplasm of amphibian skeletal muscle (250 0cm) [19]. This value seemed reasonable since the cytoplasm of the tubular cells was discontinuously compartmentalized by the cell membrane. Still, this estimate together with the spread of electronic potential could not be explained without assuming the presence

of a special electrical communication between neighboring cells. The structural basis of intercellular communications was

The cable properties of the proximal tubule

Single cable analysis In the sense that the conductive tubular fluid is enveloped by a resistive wall, the nephron has a cable-like structure. Hence,

unknown until 1970 when the frequent occurrence of gap junctions was demonstrated in the lateral cell membrane of the proximal tubules of vertebrate kidneys [20]. The gap junction is

now believed to be responsible for intercellular electric and metabolic communications in a variety of cell types [21].

cable analysis has been applied to the study of electrical According to Orci et al [22], the gap junction is seen only in the proximal tubule among renal tubular segments. Because of its frequent occurrence in the proximal tubule they referred to this initial interest was to clarify the nature of proximal tubule segment as a functional syncytium. It is obvious that, it" the cytoplasmic core is completely leakiness, one of the mysterious properties of the proximal tubule. Leakiness has been documented by many investigators covered with cell membranes having a Rm value as described [11—14], and some have referred to the proximal tubule as a above, the transmural resistance should be at least two times "leaky sieve". Our question was, if tubular cells were highly higher than the Rm value. Since the specific resistance of the permeable to Na, was their cellular potential strongly depen- tubular fluid is low (=190 - 100 0cm), the intraluminal spread of dent on extracellular Na concentration? We found firstly that clectrotonic potentials should reach far distant points showing a the cellular potential of the proximal tubule was dependent much slower spatial decay as compared to the intracellular predominantly on K concentration with little dependence on voltage attenuation, when compared at the same current Na or CL [4]. Similar results were obtained by Giebisch [3] in strength. No such spread of electrotonic potentials can be the Necturus proximal tubule. Another question arose immedi- recorded from the lumen, supporting the idea that transtubular ately. Why is the tubular wall so leaky despite the lack of cell resistance is much lower than cell membrane resistance. Previmembrane permeability to Nat? This question led us to com- ous data indicate that the transtubular resistance of the amphibpare the electrical resistances of the cell membrane and the ian proximal tubule is in the range of 97 270 0cm2 [9]. Based on these findings, we [23] initially proposed a model of the tubular wall by applying cable analysis. When we applied an electric current into cells of the proximal proximal tubule in which a vast extracellular pathway shunt tubule through a microelectrode we were surprised to observe existed within the tubular wall for ions, as illustrated in Figure that the resultant electrotonic potential spread along the tubule 2. The exact location of the shunt pathway was unknown crossing many cells (Fig. 1). The voltage attenuation along the initially, but eventually, FrOmter and Diamond [24] demontubular length was exponential, with a length constant of about strated in gallbladder epithelium that cell-to-cell junctions were 400 tm [15]. A similar voltage attenuation was found in both the site of leaky pathways for ions. Furthermore, Whittembury fluid-filled and oil-filled proximal tubules of Necturus [16]. This and Rawlins [25] demonstrated lanthanum penetration through finding indicated that tubular cells communicated electrically the tight junctions and into intercellular spaces of the proximal and that the cell membrane had a relatively high resistance tubule, giving strong evidence for the paracellular pathway when compared to the resistance of the cytoplasm. On the other leaky shunt. Ultrastructural features of the tight junction of hand, the intraluminal spread of a electrotonic potential was so leaky cpithelia, including renal proximal tubules, were characsmall in amplitude that the exact determination of the length terized by Claude and Goodenough [26] as consisting of fewer constant was difficult. To estimate cell membrane resistance, strands and discontinuous ridges or grooves. The simple cable analysis which makes use of the flat cable analysis was applied to the proximal tubule wall by simply assuming it to be a flat conductor of infinite length covered by conductor model has inherent errors in the estimate of the

properties of nephrons since the early stages of nephron electrophysiology. In a series of electrophysiological studies, our

159

Hoshi: Pitts Memorial Lecture

e

proximal tubule. A more serious problem is that the change in

the luminal membrane conductance is not reflected in an attenuation of intracellular electronic potential. Furthermore, it is important to know concomitant changes in the membrane resistance and other resistance component in the same prepa-

d4d d

)

ration in order to elucidate cellular responses to transported substances. To solve these problems, it seemed important to establish a more useful way of cable analysis which might enable us to determine not only individual cell membrane resistances, but also other resistance components by a single series of procedures. With such a purpose in mind, an analysis

of a double cable system was thought to be most desirable. We [30] and Anagnostopoulos et al [31] developed such analyses independently. Our method involves two main steps. First, current pulses Fig. 2. Hypothetical extracellular shunt pathways of the proximal tubule modeled by Hoshi and Sakai [23] before the identfi cation of the are injected into the lumen and resultant voltage attenuations Sites of leak in leaky epithelia by FrOmter and Diamond [24] and are recorded from both the tubular lumen and the cytoplasm Whittembury et al [25].

along the tubule. The recorded voltages, as a function of

distance (x), are denoted as V1(x) and V(x), respectively. The second step involves curve fitting using a computer program membrane resistance. These arise from various oversimplifica- based on the mathematical solutions of V1(x) and V(x). Mathtions. The most important source of error is the lack of ematical solutions of V(x) and V1(x) are derived according to appreciation of the radial current flow in the vicinity of the point Kirchhoff's law (Fig. 3). Double-exponential solutions were current source. Taylor [27], however, indicated that this error obtained as described below:

was not serious when the radius, or the dimensions of core

V1(x) =

.

—

aP. exp(—f3x)J/2a3(Q —

P) R110[13Q exp(—ax) cross section, was sufficiently smaller than the length constant. In order to eliminate this source of error, Anagnostopoulos and V(x) = R11,PQ[/3- exp(—ax) — a exp(—j3x)]/2af3(Q — P) Velu [281 derived mathematical solutions of voltage attenuation using Maxwell's equations. The solutions were complex and the where use of Bessel functions was needed for their application. The a = VA + V, p = VA estimated value of the combined membrane conductance (l/Ra + i/Rb) of the Necturus proximal tubule was 0.56 mmho/cm2 A = (R1/R + Ri/Ra + Rc/Ra + RCIRb)/2 (1785 11cm2) and that of the core resistivity (Re) was 11100 11cm. B = (R1/R + Ri/Ra + Re/Ra + RC/Rb)2/4

-

Guggino et al [29] also determined this resistance by using — similar but more simplified equations, and obtained a value of (RIRc/RaRb + RlR/RbR. + RiRc/RaRs 7000 — 8000 11cm for R, and 1784 11cm2 for the parallel and membrane resistance (RaRb/(Ra + Rb)). Another important source of error derives from the tubular P = —Raa2/Ri + Ra/Rs + 1 structure which may affect the flow of intracellularly injected Q = + Ra/Rs + I currents across the luminal membrane. Intracellular injection of

Ra/32/Ri

current pulses and recording of resultant voltage deflections where Ra, Rb, R, R1, R are the resistances per unit length of from inside cells failed to show changes in the luminal mem- the apical membrane, the basolateral membrane, the cytoplasbrane conductance caused by sugar- or amino acid -Nat mic core, the intraluminal fluid and the paracellular shunt. cotransport. Furthermore, a reasonable way of the calculation Similar double-exponential equations were obtained by others of individual membrane resistances (Ra and Rb) from the mean [31, 32]. By curve fitting, a set of the values for Ra, Rb, R and or parallel membrane resistance cannot be found for the free- R1 which gave the best fit to experimentally recorded V(x) and flow proximal tubule. This seems to be a serious limitation of V1(x) was determined. The experimentally observed V1(x) and the usefulness of single cable analysis of a cytoplasmic conduc- V(x) and calculated curves for V1(x) and V(x) using a set of values are shown in Figure 4. The values of these resistances tor. estimated under control conditions and in the presence of 20 The double cable model mM L-alanine in the lumen are listed in Table 1. Both experimentally recorded and calculated voltage attenuIn polarized cells, including renal tubular cells, an estimation of individual values of the apical and basal membrane resis- ations show that V(x) and V1(x) along the tubule are not tances is of particular importance since both membranes pos- parallel. This indicates that the membrane resistance divider sess different functions as well as different membrane proper- ratio, Ra/Rb, is not equal to the voltage divider ratio, (V1 — ties. Unfortunately, analysis of single cellular cable can not V)/V. This was confirmed by Cook and Frömter [32]. The yield individual values of the luminal and peritubular membrane resistance divider ratio is equal to the ratio of the integrals of resistances. Although the voltage divider ratio has frequently V1(x) — V(x) to that of V(x). In other words, Ra/Rb is equal to been used to determine the ratio of Ra/Rb, the relationship of the ratio of the area surrounded by the two curves for V(x) and the Ra/R, to the voltage divider ratio is not so simple in the V1(x) to the area encompassed by the curve for V(x). This can

Hoshi: Pitts Memorial Lecture

160

V,(x +

V1(x)

R1

Fig. 3. Current flow in a double cable model of the proximal tubule in the case of intraluminal current injection. Abbreviations are: Ra, the apical membrane resistance; Rb, the basolateral membrane resistance; R, the paracellular shunt resistance; R0 the cytoplasmic core resistance; R1, the resistance of the luminal fluid; 0, outside medium; C, cytoplasm; L, the lumen; 10 is the total current injected into the lumen. V(x) and V4(x) are the voltages as a function of x in the cytoplasmic conducting core and the tubular lumen, respectively. Used with permission

x)

from [30].

40 30

20

E w

10

>0

8

6 Fig. 4. Experimentally recorded V,(x) and V(x) under control conditions (0) and in the

presence of 20 m alanine in the luminal fluid (x). Current pulses of a constant strength (5 x A) were applied into the lumen. Each

4

l0 point is the mean of data from S different

0

0.2

0.4

0.6

0.8

1.0

Distance, mm

Table 1. Values of Ra, Rb, R and R0 of Triturus proximal tubule determined by double cable analysis under control conditions and in the presence of 20 mi L-alanine in the luminal perfusion solution Ra Conditions

Control

R,

Rb

1350

2100

166

800

24

4200 200

Data are presented as means SEM, N =

150 1

5,

decreases significantly when the luminal membrane is exposed

to alanine. In contrast, the intraluminal attenuation curve, V1(x), changed only slightly. V1(x) is determined primarily by the low resistance extracellular shunt. Quantitatively, the con-

3400

ductance change in the luminal membrane caused by 20 m alanine was estimated to be 0.51 mS/cm2. In contrast, the

3

In the presence of 20 mri alanine

V1(x). This indicates that the luminal membrane resistance

R0

fkm

Clcm2

12

animals. The lines were drawn by computeraided calculations based on the mathematical solutions of V1(x) and V(x) and using a set of resistance values which gave the best fit to the data. Used with permission from [301.

3900 900

Used with permission

from [30].

basolateral membrane resistance increased significantly. Lang,

Messner and Rehwald [101 observed, in the frog proximal

tubule, that the transfer number of K for the basolateral membrane decreased rapidly from 0.6 to 0.2 when 10 mM phenylalanine was injected into the lumen. This reduction was ascribed to the property of the inwardly rectifying K conduc-

be understood when we consider that all current which enters the cells through the luminal membrane leaves the cells through the basolateral membrane. Without taking this into account, the resistance divider ratio (RaIRb) is significantly overestimated. Exposure of the luminal membrane to 20 mM alanine caused a marked shift of the curve for V(x) approaching the line for

tance of the basolateral membrane and to cell membrane depolarization. In the intestinal epithelium, a prolonged exposure to an actively transported sugar or amino acid was shown to result in a gradual increase in basolateral membrane conductance [331. Although Lang et al found a similar tendency in the frog proximal tubule after prolonged exposure to phenylalanine,

161

Hoshi: Pius Memorial Lecture

such a change was not seen clearly in the Triturus kidney. Data

—10

exposure of the mucosal surface to L-alanine or D-glucose

junctions) are not affected by Nat-sugar or Nat-amino acid cotransport. Even at a higher concentration (40 mM), alanine caused a decrease in the transtubular resistance by only 15%

,1

—55 E

d —60

—65

—70

[36].

A multiple cable model was recently proposed by Weber and Frömter [37] to analyze the behavior of the lateral space. In this

0

Ett

—5

changed only slightly (by about 10%) by 20 mM alanine. This is in sharp contrast to the findings in the small intestine, in which

induces 200 - 300% increase in transmural conductance [34]. In the small intestine, the intercellular space expands considerably and the ultrastructure of the tight junction also changes under absorptive conditions [35]. However, no such expansion has been seen in the proximal tubule. A minor change in R suggests that the intercellular communication pathways (probably gap

-

0

derived from double cable analysis also show that the shunt resistance (R) and the cytoplasmic core resistance (Re) are

I

—75

model, the lateral space was modeled as a series of n cable which run in parallel with and connected to the lumen and

Elm

j

} Ebm

I

Ebm

5. Electrical potential across the proximal tubule of Triturus cellular cable. The mathematical solution of this model led to n Fig. kidney before (—) and after (---) addition of D-glucose in the luminal

+ 2 exponential functions. Due to the complexity of these

equations, direct use of the solutions in data fitting is difficult.

fluid. Ebm is the transmembrane potential across the basolateral mem-

brane and Ejm the transmembrane potential across the luminal mem-

However, they could show that the influence of the lateral brane.

is the change in the transtubular potential and Ebm the

space on the overall cable property was insignificant in Nectu-

change in the basolateral transmembrane potential induced by intralu-

rus proximal tubule under control conditions. Whether the

permission from [43].

lateral space changes during sugar or amino acid transport has not been analyzed.

minal injection of D-glucose. PD, potential difference. Used with

abolished by addition of phlorizin. The amplitude of the membrane depolarization as well as that of transtubular potential change obeyed Michaelis-Menten kinetics. The generation of transport such electrical changes was absolutely dependent on the presSugars ence of Na in the perfusion solution. Similar electrical changes The pioneering work by Crane [38, 39] on Nat-coupled uphill were observed in the small intestine [44] and in the rat proximal transport of sugars and a subsequent demonstration of electro- tubule [45]. The specificity of the electrogenic glucose carrier can be genicity of active sugar transport in the small intestine [40—42] led us to a series of investigations on the effects of sugars and assessed by comparing K values of various monosaccharides. amino acids on the cellular potential of the proximal tubule. These sugars are perfused at various concentrations and the This line of study seemed particularly important for understand- electrical response is recorded. D-glucose and a-metyl-D-gluing the role of a constant Na concentration within the tubular coside caused depolarization in all tubules tested. An unexlumen which is related to the leakiness of the proximal tubule. pected observation was that D-galactose did not induce depoObservations on the effect of D-glucose indicated that the larization, although this hexose caused an increase of mucosal peritubular transmembrane potential depolarized by about 15 negativity in the Triturus intestine like in other vertebrate mY at maximum, and the luminal side become more negative, intestines. Also, two types of proximal tubules could be distinby I - 1.5 mY, when 10 m or a more concentrated D-glucose guished in the Triturus kidney; one responded to D-mannose was introduced into the lumen [43]. The resultant changes in the with depolarization but not to D-xylose, while the other repotential profile of the tubular cells are shown in Figure 5. The sponded to D-xylose but not to D-mannose [46]. Silverman, profile change can be interpreted as the result of generation of Aganon and Chinard [47] showed two types of sugar carriers in an electromotive force at the luminal membrane upon exposure the luminal membrane of dog proximal tubule which were called to D-glucose. The resultant current circulates through two G- and M-types. Our results together with Silverman's findings separate pathways. One is via the luminal membrane, and the suggest that renal tubular sugar carriers differ in specificity from other is a circuit formed by the peritubular membrane and the intestinal sugar carriers. For the intestinal sugar carrier, the paracellular pathway. Consequently, both the luminal and importance of C3-OH and C4-OH differs significantly among peritubular membranes depolarize and the luminal side be- vertebrate species [48]. A more detailed study of the renal and comes more negative. Owing to a much higher resistance of the intestinal sugar carriers in the same animal species is needed for peritubular membrane than the shunt resistance, the amplitude a more precise characterization of the specificity differences. of the membrane depolarization is much greater than that of the Amino acids transtubular potential change. The ratio of the two electrical In the Triturus proximal tubule, the Nat-dependent mempotential changes (LWb!LWtr) is equal to the ratio, R1,/Rs. The glucose-induced electrical changes were found to be promptly brane depolarization is seen only for neutral amino acids [49, Electrophysiological observations on sugar and amino acid

162

Hoshi: Pitts Memorial Lecture 25

20

Alanine 10mM

E

Lysine 10mM

Aspartate 10mM

CD

C

15

CD

C.) CD

C

10

CD

0 0.

•0 CD

0

5 Li

0 CD

0 C

0

Tris Mannit.

Na Na

Li

Tris Mannit.

E —5

—10

Li

Tris

Fig. 6. Average amplitudes of intracellular voltage changes induced by intraluminal applications of three different amino acids at 10 mM in the presence and absence of Na. The evoked potentials were recorded across the peritubular membrane of Triturus proximal tubule. Abbreviations are: Na, normal Ringer's solution; Li, Li-Cl-substituted medium; Tris, Tris-Cl-substituted medium; mannit, mannitol-substituted medium. Positive potential change means the depolarization, and negative change the hyperpolarization. Note a sizable hyperpolarization is induced by alanine in the absence of Na except for the Li-medium. Used with permission from [49].

50]. Basic amino acids cause a marked depolarization of the Thus, lysine transport across the luminal membrane may be an cellular potential, but depolarization is found to be entirely electrogenic uniport. On the other hand, lysine exit across the

independent of Na (Fig. 6). Acidic amino acids, such as basolateral membrane appears to be Na dependent. In a aspartate, cause no electrical change although tissue uptake is Tris-substituted medium, lysine-induced depolarization beabsolutely Na dependent [50]. These findings in Triturus differ comes smaller with time when intraluminal pulse injection tests from those in the rat, in which all the three types of amino acids are repeated. After extinction of the responsiveness of the cell

induce cell membrane depolarization in a Na dependent to lysine, addition of Na to the peritubular fluid causes a manner [51, 52]. gradual restoration of the response. Furthermore, Na influx The contradictory findings on the Na dependence of the across the peritubular membrane was significantly greater in the basic amino acid-induced depolarization in Triturus and rat kidneys have not been explained except in terms of species difference. Contradictory results have also been obtained in membrane vesicle experiments. Busse [53] and Hilden and Sacktor [54] showed the Na dependent nature of arginine transport by demonstrating a slight enhancement of uptake by Na as compared to K in the presence of a inside-negative membrane potential. However, these investigators showed that the uptake was usually large in a Nat-free mannitol medium. Recent studies by Hammerman [55] and McNamara, Rea and

Sega! [56] failed to show a Na dependence for lysine or arginine transport by brush border membrane vesicles. These studies, however, showed a clear voltage-dependence of the transport. As to the transcellular transport, Ullrich, Rumrich

and Kloss [57] demonstrated that active transport of basic amino acids was Na dependent as shown by formation of a zero-flux concentration difference across the tubule only in the

presence of Na. Our recent reinvestigation in the Triturus nephron confirmed

the Na independence of the lysine-induced depolarization.

The investigation also ruled out the possibility of the H dependence which was demonstrated in the mullet kidney by Lee and Pritchard [58]. In addition, we gained new information

suggesting a role for Na in basic amino acid transport. Lysine-induced depolarization obeys Michaelis-Menten kinetics regardless of the presence or absence of Na and strongly depends on the initial level of the transmembrane potential.

tubule which had been preloaded with lysine when examined before and after the addition of Na with ouabain (0.5 mM) to the peritubular fluid (Nunokawa and Hoshi, unpublished observation). These findings indicate that Na is needed for the exit process across the peritubular membrane. Our findings support a possible mechanism of basic amino acid transport which was expected but not supported experimentally by Shalhoub et al [59] and Samarzija and FrOmter [51]. The mechanism is that basic amino acids enter cells voltage-dependently without interacting with other cations, and depart cells through a NaI amino acid exchange mechanism to overcome the adverse voltage gradient across the peritubular membrane (Fig. 7). Extinction of responsiveness of basic amino acids in a Na-free medium and restoration of the response by peritubular Na are more clearly observed in a Tris-substituted medium than in a mannitol-substituted medium. The substituting cation may affect the cell responsiveness to basic amino acids. Although the mechanism of a more rapid inhibition by Tris is unknown, this may explain the discrepancy, at least in part, regarding the Na dependence of luminal membrane depolarization. Other evidence supporting the uncoupled uniport across the luminal membrane is the unchanged voltage dependence of the lysineinduced deporarization before and after Na replacement. The slope of voltage dependence is unchanged, suggesting unchanged charge transfer [36] (Fig. 8). Controversial results regarding the electrogenicity of acidic amino acid transport have not been explained, except to be

Hoshi: Pius Memorial Lecture

163

gated channel model of the Patlak-Lauger type [69] seems Lys *

Lys Na *

favorable for explaining a high energy barrier phenomenon and tightly coupled transfer of the cotransported solutes through a

channel-like structure. These models, however, appear to be insufficient to explain some electrophysiological aspects of cotransport, particularly electrical phenomena in intact tubular cells. As described above, the luminal membrane conductance increases when a sugar and a neutral amino acid are cotransported with Nat In intact tubular cells, the increased conductance returns to the original state immediately when the sub-

strate is flushed even after a prolonged exposure to the substrate (Fig. 9). This indicates the existence of a gate-opening mechanism for Na associated with or located in the vicinity of the substrate binding site to which the substance can bind only Fig. 7. A possible mechanism of lysine transport across the luminal and basolateral membranes.

from the outside medium. Even in the absence of Na in the outer medium, the substrate binding from the medium appears

to be able to induce the gate opening. In such a case, the resultant increase in Na conductance results in a hyperpolarization of the membrane (the alanine-induced hyperpolarization, Fig. 6). The reversal potential phenomenon is also seen in of 1:1 coupling with Nat, since the kinetic effects of Na the voltage response. The depolarization under normal condiconcentration on cellular uptake of aspartate were explained by tions vanishes when the membrane potential decreases to a a 1:1 coupling cotransport kinetic model [50]. However, recent certain level, and the response is converted to hyperpolarizainvestigations on membrane vesicle uptake show a more com- tion when the transmembrane potential is made more positive plex interaction of the carrier not only with Na, but also with beyond the null response level [70] (Fig. 10). This finding H in the cotransport step and with K (or H) in the antiport indicates that the transmembrane potential shifts toward an step [60]. The net charge transfer is dependent on the stoichi- equilibrium potential for the cotransport when the gate is ometry of charged species in co- and antiport steps. As sug- opened. Therefore, once the gate has opened, Na can flow gested by Heinz, Sommerfeld and Kinne [61], the stoichiometry through the carrier in a symmetric manner as demonstrated by may be variable, depending on the experimental conditions. Jauch, Petersen and Lauger [71] in whole cell recordings of Species differences may also exist in the coupling stoichiom- pancreatic acinar cells. These electrophysiological findings can be explained using etry. Thus, the modes of interaction with ions are much more the following model [71]. As illustrated in Figure 11, the complicated in amino acid transport than initially expected. cotransport processes can be divided into three steps. The first Recently, transport of /3-amino acids, such as taurine and is a stereospecific binding of the substrate to its own carrier site f3-alanine, has been shown to involve an interaction with (referred to as "recognition" in Fig. 11). The second is the multiple Na and one Cl [621. In addition, purely proton- immediate opening of the gate for the Na pathway which is dependent processes exist in the luminal membrane, such as, probably due to a conformational change of a portion of the gate the H/dipeptide cotransport [63, 64]. The H-dependent pep- adjacent to the substrate binding site (the gate mechanism). As tide transport is also Na-dependent in the sense that its driving a result, the carrier protein becomes conductive to Nat, and force, that is, the H gradient across the membrane, is formed Na flow occurs along its electrochemical potential gradient. by the Na/H antiport. In the small intestine, H gradient The third is the coupling of substrate flow to Na flow within formation is totally dependent on the Na/H antiport [65], the carrier molecule when the Na flow occurs in the normal while in the renal proximal tubule, not only the Na/H direction. Unfortunately, the mechanism of flow coupling is antiport but also a ATP-driven H pump [66] may participate in unknown, but a tight coupling mechanism must be considered the gradient formation, although the relative importance of the in this process because most experiments have shown a fixed latter is not known. (unchanged) stoichiometry. The model in Figure 11 implies that, under intact cellular Possible mechanism of operation of the cotransporters conditions, the carrier exhibits a strong rectification. Such Although the amino acid sequence for the Na/D-glucose conditions are favorable for vectorial transport from the lumen based on species difference. Nonelectrogenic transport of Laspartate in the Triturus kidney was simply interpreted in terms

cotransporter has been determined [67], the molecular structure

to the intracellular space at the carrier level, even in the

and location of the binding sites for Na and the substrate are situation of a progressive increase of intracellular substrate largely unknown. Accordingly, it is premature to show a model concentration. In membrane vesicles, the carrier behaves more which explains the mode of operation of the cotransporters. symmetrically. Further studies in single cells under varied Nevertheless, some current models are useful for explaining conditions are needed in order to understand the behavior of the major characteristics of cotransport processes. For example, an cotransporter under variable experimental conditions. One interesting problem related to the carrier operation is the asymmetric gated channel (or pore) model proposed by Kessler and Semenza [68] explains not only most kinetic data obtained voltage-dependence of sugar transport by the phlorizin-sensiin membrane vesicle experiments but also effects of proteases tive sugar carrier in the absence of Nat Hilden and Sacktor or mercurials acting from inside the membrane vesicles. A [73] demonstrated that Na-independent sugar uptake by renal

164

Hoshi: Pills Memorial Lecture

A Normal medium 20

B Na-free medium 20 C

C

0

IN 15

15

(0

N

(0

(0

0 0.

00.

Fig. 8. Voltage-dependence of the lysine-evoked depolarization in the presence (A) and absence (B) of Na in the Triturus proximal tubule. The potential changes evoked were recorded under conditions of spontaneous transmembrane potential (•), in the presence of a depolarizing current (0), and in the presence of a hyperpolanzing current (x). The depolarizing and hyperpolarizing currents were passed intracellularly from a point 100 — 200 sm apart from the recording site. Used with permission

. 10

10

a) a)

0 > a)

5

a)

C U)

0

20

40

60

80

100

Transmembrane potential, mV

A

0

20

40

60

80

100

Transmembrane potential, mV

from [36].

B

0 Ala. 20 mM

Ala. 20 mM

p

I

0

Ala. 20 mM__/

Fig. 9. The alanine-induced depolarization of the basolateral membrane and concomitant changes in the resistance divider ratio in the Triturus proximal tubule. The luminal membrane was exposed to 20 mM alanine during the periods indicated by a horizontal line. Cut-rent pulses (1O— A) were injected through a microelectrode into the lumen at 75

10 mV

1 mm

brush border membrane vesicles was enhanced significantly by inside-negative transmembrane voltage. They interpreted such voltage dependence in terms of a negatively charged state of the carriers and their voltage dependent recycling (or reorientation) within the membrane. However, our observations showed that

the Nat-independent voltage-dependent sugar transport by intestinal and renal brush border membrane vesicles were enhanced significantly by a lowering of external pH [74]. Furthermore, in the absence of Na, sugar transport potential across intestinal epithelium was also augmented by lowering the pH of the medium. And, the augmented transport potential was abolished by phlorizin. This indicates that H can substitute for

the role of Nat In addition, the gate for Na becomes conductive to H following substrate binding, and the flow of H can couple with sugar flow as in the case of Nat-glucose

cotransport. Such an ability of H to substitute for Na, however, is seen only in the NafD-glucose cotransporter. This does not apply to the Na/neutral amino acid cotransporter.

tm apart from the recording site. An increase in amplitude of pulses indicates a relative increase in the luminal membrane conductance with respect to the basolateral membrane conductance. A. Short-period exposures; B. A prolonged exposure to alanine. Interrupted recording lasted 10 mm. Note that a prolonged exposure does not after the feature of abrupt recovery of the luminal membrane resistance upon flushing the intraluminally added alanine. Used with permission from [72].

Kinetic parameters for sugar and amino acids and an optimality principle in the distribution of the transport systems in the proximal tubule Electrophysiological techniques combined with intraluminal microperfusions with varied solutions enable us to determine the kinetic parameters of transport with relative ease. Additionally, the K values estimated by electrical recordings can be compared with those estimated by flux measurements. In the case of Triturus kidney, a transported substance with its tracer added to the peritubular fluid is rapidly introduced into the lumen by virtue of continuous nephrostome activity, and is taken up by the cells through the luminal membrane processes [48]. The K value for alanine, determined by electrical and flux measurements were found to be identical. Based on this finding, the values of the maximum influx (max) for other substances can be calculated from the values of PDmax by referring to the relationship between the values of 4PDmax and JmaX for alanine.

Hoshi: PUts Memorial Lecture + 20

Glucose

Glucose

15mM

15mM

165 Membrane

0 Cotransport carrier

E

a C

a

00. a C

a

.0 E

a

Case 1

Case 2

Fig. 10. Effect of initial level of the transmembrane potential on the glucose-induced depolarization in the Triturus proximal tubule, The

level of transmembrane potential was varied by injecting constant current of various strength into a neighboring cell. Note that there is a null-response level at around —10 —15 mY and hyperpolarizing responses are induced at more positive levels beyond the null-response level. Used with permission from [70].

Fig. 11. A model of NaIneutral organic solute cotransporters which explain electrophysiological findings associated with Na/sugar or Na'Vneutral amino acid cotransport in intact cells ofTriturus proximal tubule. Used with permission from [721.

For a substance which is transported by a nonelectrogenic mechanism, such as aspartate, the kinetic parameters are

5

determined by flux measurements only. Plotting the max value

against K value for individual substances reveals a definite relationship between these two parameters; the lower the K

Alanine 0

4

value, the lower the 1max value [34] (Fig. 12). This indicates that

Aspartate

a high affinity system has a low capacity and vice versa. The regression line relating max to K can be expressed as

>

'max = 0.45 K' + 0.35

a a>

where 'max and K' are relative values to the respective values for D-glucose. Surprisingly, a similar relationship is seen in the

rat proximal tubule for amino acid transport when kinetic

E

0

3

a > a

2

a cC

Lysine

parameters for various amino acids, summarized by Silbernagl [75], are plotted in the same manner [761. In the case of the rat

kidney, the regression line is given as 'max 0.44K' + 0,94 when the value for L-glutamate is taken as unity and relative values are plotted. This suggests that the numerical relationship has a physiological meaning. To clarify this relationship, a simulation study was carried out using a mathematical model of a flow reactor (absorption) system simulating Triturus proximal tubule [76, 77]. It was assumed that a solution containing a transported solute at an initial concentration of C0 flowed into the tubule at an initial flow rate, F0. The solute was absorbed by interacting with its carriers which were distributed evenly on the inner surface of the system. The transport was assumed to follow MichaelisMenten kinetics. For the sake of simplicity, approximately half of the fluid was regarded to be absorbed at a constant rate during passage through the entire length of the system, and the backleak of the transported solute was assumed to be negligible. Computer-aided calculations of the intraluminal concentration as a function of x (the longitudinal axis) yielded a concen-

2

4

6

8

10

Relative value of K,

Fig. 12. Relationship between K, and V,,,

values for five different substances determined in Triturus proximal tubule. Relative values of K, and Vma with respect to respective values for glucose are plotted. Modified from [72] with permission.

tration profile resembling that observed by micropuncture studies. For example, when using the parameters for D-glucose, glucose concentration rapidly decreased in the early half of the system to reach nearly zero at about the middle of the system,

when C0 was set at 3 m and F0 15 nllmin. An increase in the K, without changing other parameters

results in a slower decrease in concentration, whereas an

Hoshi: Pitts Memorial Lecture

166

3

Fig. 13. Simulated concentration profile for actively transported sugar or amino acids and effects of perturbation of parameters of transport kinetics (K1 or V,,,) on the profile.

E

0

2

a)

a 1.) C 0 0 a)

C E

Ca

C

1

2

3

4

5

10

6

Distance, mm

A mathematical model was constructed by using geometrical and functional parameters for Triturus proximal tubule. The standard profile (a heavy line) was obtained by using K1 and Vmax values for glucose and assuming its initial concentration of 3 mat. Initial flow rate was set at 15 nl/min and half of the fluid flowed into the tubule was assumed to be absorbed during passage through the entire length of the tubule. The deviation of the perturbed profile from the standard one resulted from an arbitrary increase in K, or 'max value is expressed by an absolute error integral as indicated by shaded area.

0.09

independent increase in max causes a faster drop in concentra-

tion as expected from their kinetic effects. Fixing the K, at a high level, stepwise increases in the max brings about a gradual approach of the concentration profile towards the original, then an over-shift beyond the standard profile (Fig. 13). Accordingly,

0.08

when the extent of deviation from the standard curve was quantified by the absolute value of area surrounded by both the

standard and a perturbed curves and plotted against relative increases in max, a downward parabolic curve was obtained (Fig. 14). Thus, the value of max' which gives the minimum E

mined graphically is exactly the same as that calculated by using the equation of regression lines. This indicates that the

a)

renal tubule which reabsorb, within a limited length, a wide variety of physiologically important substances exhibiting diverse affinities. Moreover, it should be pointed out that the

mode of compensation is in accord with the principle of minimum error (or the optimality principle) which is one of most important and common principles in system engineering.

C(x)Idx

0.05

0 0.04

This is probably due to a higher density of carriers, so as to reabsorb its substrate at an efficiency nearly the same as that of a high affinity system. This seems an ingenious device in the

—

0.06

value of deviation, can be determined. Surprisingly, max deterlow affinity system is compensated by the higher value of

Absolute error area = fo1IC(x)

0.07

<

0.03

0.02

0.01

Low resistance K-seIective pathways in the early distal tubule Electrophysiologically, the early distal tubule (eDT) of the

amphibian kidney can be characterized by: a lumen-positive transtubular potential; a nonelectrogenic luminal membrane

transport of Na and Cl- coupled with K (2Cl/Na/K

— 1

2

3

4

5

6

7

Relative value of V5 14. Effect of increase in V,,,, value on the absolute error integral Fig. between the standard profile and the profile for a low affinity system with a K, value 5 times higher than that of the standard profile. Note that the minimum value of the error integral is seen at a certain value of "max (Unpublished data.)

cotransport); a net outward transfer of negative charge; and, a high sensitivity of these electrical properties to furosemide [78, 79]. The same properties are seen in the Triturus eDT [80]. These characteristics are the same as those in the mammalian thick ascending limb of Henle as reviewed by Greger [81]. Kinetically, effects of intraluminal Nat, K or C1 concentration on the equivalent short circuit current are similar to those absence of either of these three ions suppresses completely. in the mammalian thick ascending limb of Henle. As shown in The K1 value for Na is 2.6 m in the presence of 106 mrs'i Cl Figure 15, the short circuit current (I) is absolutely dependent and 2.5 ma's K, whereas that for K is 0.9 ma's in the presence

on the presence of Na, K and Cl in the luminal fluid. The

of 106 ma's Cl and 5 mM Nat. After testing several kinetic

i*

A Na-dependence

+30 + 20

> +10 0 Na

0 3

3

0 3

5 3

0 3

106

106

106

106

5

::: K*

C1 106

167

Hoshi: Pitts Memorial Lecture A 0.4

30mM K*

0

2

30mM K*

4 6 8 10

Na concentration. m

B K-dependence

Na

K 00 Cl- 106

3mM K

>

C 0 Na 50

K

3

mm

B

0 5

0 0

3 5

0 5

106

106

106

106

ftt

3mM K

110 mV 1

C C1-dependence + 30

3mM K

100

0

100

50

50

50

0 50

3

3

3

3

30 mM Na*

K concentration, mM

30 mM Na*

3mM Ba +

30 mM Na

0.4 0

3mMNa

3 mM Na*

-.-.----

3mMNa

10 mV

0 20 40 60 80 100 C1 concentration, m

Fig. 15. Dependence of the lumen-positive transtubular potential (V,)

and the equivalent short circuit current (Ifl,) on Na (A), K (B), and C1 (C) concentrations. Triturus early distal tubule. The lumen of the tubule was perfused from a Bowman's capsule with a solution of varied composition as indicated below each potential tracing. Dependence of

I, on individual ions was studied in the presence of 106 m C1 and 3 mM K for Na, in the presence of 106 mss C1 and 5 mri Na for K, and in the presence of 50 mas Na and 3 mM K for C1. Used with permission from [82),

1 mm

Fig. 16. Effects of Ba2 (3mM) on spiky diffusion potentials induced by tenfold changes in intraluminal concentrations of K (A) and Na (B).

In order to suppress the 2CF/Na/K cotransport, Na was totally replaced with Tris in A and lO M furosemide was added to the perfusion solutions in B. Used with permission from [83].

potential of the similar (spiky) shape, but the amplitude is much

less than that of the K diffusion potential, indicating the

permeability sequence of k >> Na > P1. So far, the

equivalent electric circuit of eDT has not been established models, I could be described by a single equation as a function because of difficulties in recording stable intracellular potentials. Our preliminary observations showed that both the inof Nat, K and C1 concentrations [821. Of particular interest among electrical properties of this traluminal and intracellular voltage attenuation curves for insegment is the low resistance of the tubular wall. In Triturus, traluminal current injection were parallel, suggesting a lack of the transtubular resistance is 31 flcm2 [83] or 57 11cm2 [84], intercellular electric communications. Furthermore, the memwhereas it is 9.1 — 15 11cm2 in Ambystoma and Amphiuma [78]. brane resistance does not seem to be exceptionally low since These values are much lower than the transtubular resistances the input resistance measured with an intracellularly located of the proximal tubules of the same animals. In addition, the microelectrode was 17.5 M11 on the average [831. These findtranstubular resistance is much lower than the sum of Ra and ings, together with observtions by Guggino, Oberleithner and Rb, suggesting the presence of a low resistance extracellular Giebisch [79], strongly suggest that the low resistance pathway shunt in the epithelium [79]. Observations of diffusion potential is extracellular.

across the tubule strongly suggest that there is a highly Kselective diffusion path in the wall. This is based on the finding that a ten-fold change in K concentration generates a transient (spiky) diffusion potential reaching a peak height of 58 mV [83] (Fig. 16). This value is a maximum amplitude of the diffusion

potential through a single barrier. Replacement of Cl— with SO converted the shape of the diffusion potential to a plateau type, indicating that the path has a finite C1 permeability. A ten-fold change in Na1 concentration produces a diffusion

It is interesting to note that the principal function of this segment is to reabsorb Na and Cl, and that this function is absolutely dependent on the presence of K in the lumen. This situation is analogous to that of the proximal tubule in which the

principal transport functions are Na dependent and a high enough concentration is maintained in the lumen by virtue of the presence of the extracellular shunt. In support of this idea, intraluminal K concentration does not change significantly even when cell conditions are modified by ouabain or other

Hoshi: Pitts Memorial Lecture

168 A

B

1 mm

NaCI 5

0 5

Tris-CI 90

90

KCI 2.5

2.5 0

5 90

5 90

90

Fig. 17. Changes in the transtubular potential of Triturus early distal tubule induced by changes in ionic composition of the intraluminal peifusion solution (indicated in mM below the tracings) and the cessation of forced perfusion with a K-free solution. The period of stopped perfusion is indicated by a horizontal bar. A. In the absence of Ba2, B. In the presence of 2 m Ba24 during perfusion with the K-free solution. Note that besides complete suppression of generation of lumen-positive transtubular potential during the stopped perfusion, spiky diffusion potentials due to the change in K concentration is suppressed. This may probably be the result of remaining effect of Ba2. The tracing is a part of repeated

2

experiments. Used with permission from [831.

1 mm

KCI

2.5

NaCI 5

Iris-Cl 90 BaCI2 0

0 5 90 2

drugs [79]. In addition, a rapid K4 movement into the lumen is

2.5 0 5 90 0

5

Reprint requests to Takeshi Hoshi, M.D., Laboratory of Physiology,

seen when a forced perfusion of the lumen with a K4-free School of Food and Nutritional Sciences, University of Shizuoka, solution is stopped. Under such conditions, the suppressed Shizuoka 422, Japan. lumen-positive transtubular potential (or outward transfer of negative charge) recovers rapidly, as shown in Figure 17 [83]. The K entry may occur via both the luminal membrane and the paracellular pathway. Schwab and Oberleithner [85] and Guggino et al [79] concluded that participation of the two pathways might be 1:1 on the basis of assumption that Ba2 inhibits only the cell membrane K pathway. However, as shown in Figure 16, Ba2 strongly inhibits not only the K diffusion potential, but also the Na diffusion potential, suggesting that the K-selective diffusion path is also sensitive to Ba2. Consideration of these observations leads us to conclude that the role of the shunt in the regulation of intraluminal K4 must be greater than currently believed. Further studies on K movements under various conditions as well as the establishment of the equivalent electric circuit of eDT will clarify not only its physiological significance, but also the specific role of the extracellular shunt of this segment. This may also contribute to a further understanding of the functional and structural integration of the nephron. Acknowledgments I wish to express my thanks to the International Renal Commission for the honor of receiving the Robert F. Pitts Memorial Lectureship Award and for the opportunity of presenting this address. Much of the

References 1. Prrrs RF: Physiology of the Kidney and Body Fluids. Chicago, Year Book Medical Publishers, 1963 2. WINDHAGAR EE, GIEBIscH G: Comparison of short circuit current and net water movement in single perfused proximal tubules of rat kidneys. Nature 191:1205—1207, 1961

3. GIEBISCH G: Measurement of electrical potential difference on single nephrons of the perfused Necturus kidney. J Gen Physiol 44:659—678, 1961

4. SAKAI F, HosHi T, HAGA M, ENOMOTO Y: Membranpotential an

der Nierentubuli des Triturus pyrrhogaster. Jpn J Pharmacol 11:65—71, 1961

5. Hosu T, SAKAI F, HAGA M: Electrical properties of renal tubules of the newt. J Physiol Soc Japan 24:378, 1962 6. FROMTER E, MUELLER CW, WICK T: Permeability properties of proximal tubular epithelium of the rat kidney studied with electrophysiological methods, in Electrophysiology of Epithelial Cells, edited by G GIEBISCH, Stuttgart, Schattauer, 1971, pp. 119—146 7. GIEBISCH G, WINDHAGER EE: Electrolyte transport across renal tubular membranes, in Handbook of Physiology, Section 8, Renal Physiology, edited by J ORLOFF, RW BERLINER, SR GEIGER, Am Physiol Soc, Washington DC, 1973, pp. 315—376 8. BOULPAEP EL, GIEBISCH G: Electrophysiological measurements on the renal tubule, in Methods in Pharmacology Vol. 4B, edited by M MARTINEZ-MALDONADO, New York, Plenum, 1978, pp. 165—193

9. BOULPAEP EL: Electrical phenomena in the nephron, Kidney mt

work discussed in this lecture was product of joint efforts with my 9:88—120, 1976 colleagues at the University of Tokyo and at Tohoku University. I 10, LANG F, MESSNER G, RERWALD W: Electrophysiology of sodiumshould like, in particular, to acknowledge my great debt to Dr. K. coupled transport in proximal renal tubule. Am J Physiol 250: Matsuda, Professor Emeritus, Tohoku University, who taught me F953—F962, 1986 electrophysiology during my graduate course, and to Dr. F. Sakai, 11. Hosrno T, SWANSON RE, VISSCHER MB: Excretion of Na22 and Professor Emeritus, the University of Tokyo, with whom our electrophysiological study of nephron was initiated. Dr. Sakai taught me the

usefulness of Triturus in single nephron studies. Also I thank Dr. George Greeley, a Guest Professor at the University of Shizuoka from the University of Texas Medical Branch, Galveston, who revised the manuscript. My debt is also due to Mrs. Miyuki Shimizu, who assisted

in the preparation of the manuscript. Most of our studies were sup. ported by the Grants-in-Aid from the Ministry of Education, Science and Culture of Japan.

K42 by the perfused bullfrog kidney and the effect of some poisons. Am J Physiol 184:542, 1959 12. WILDE WS, MALVIN RL: Movement of peritubular Na across the nephron wall during stop-flow. Fed Proc 18:170, 1959 13. PENE JC, MALVIN RL: Transcellular diffusion of non-electrolytes across the renal tubular epithelium. J Gen Physiol 45:643, 1962 14. OKEN DE, WHITTEMBURY G, WINDHAGER EE, SCHATZMANN HJ, SOLOMON AK: Active transport of sodium by the proximal tubule of Necturus. J Clin Invest 38:1092, 1959

169

Hoshi: Pitts Memorial Lecture 15. HOSHI T, SAKAI F: A comparison of the electrical resistances of the

The interaction between active sodium and active sugar transport.

surface cell membrane and cellular wall in the proximal tubule of the newt kidney. Jap J Physiol 17:627—637, 1967

41. BARRY RJC, DICKSTEIN 5, MATHEWS J, SMYTH DH, WRIGHT EM:

J Gen Physiol 47:1043—1058, 1964

Electrical potentials associated with intestinal sugar transfer. J

16. WINDHAGER EE, BOULPAEP EL, GIEBSCH G: Electrophysiological

studies on single Nephrons, in Proc III Intern Congr Nephrol, (Vol. 1) Physiology, edited by JS HANDLER, Basel, Karger, 1967, pp. 35—47 17.

Physiol 171:316—338, 1964

42. LYON I, CRANE RK: Studies on transmural potential in vitro in relation to intestinal absorption. I. Apparent Michaelis constants

HODGKIN AL, RUSHTON WAH: The electrical constants of a

crustacean nerve fibre. Proc Roy Soc London B 133:444—479, 1946 18. WEIDMANN S: Electrophysiologie der Herzmuskelfaster, Bern, Hans Huber, 1957

19. FArr P, KATZ B: An analysis of the end-plate potential recorded with an intracellular electrode. J Physiol 115:320—370, 1951 20. SILVERBLATT FJ, BULGER RE: Gap junctions occur in vertebrate renal proximal tubule cells. J Cell Biol 47:513—515, 1970 21. GILULA NB, REEVES OR, STEINBACH A: Metabolic coupling, ionic coupling and cell contacts. Nature 235:262—265, 1972 22. ORCI L, HUMBERT F, BROWN D, PERRELET A: Membrane ultrastructure in urinary tubules. Jnt Rev Cytol 73:183—242, 1981

23. H0sHI T, SAKAI F: Measurement of electrolyte permeability, transmembrane and transtubular resistance in proximal tubules. J Japan Med Assoc 55:708—713, 1966 (in Japanese) 24. FROMTER E, DIAMOND JM: Route of passive ion permeation in epithelia. Nature New Biol 235:9—13, 1972 25. WHITTEMBURY G, RAWLINS FA: Evidence for a paracellular pathway for ion flow in the kidney proximal tubule: Electronmicroscopic demonstration of lanthanum precipitate in the tight junction. Pflugers Arch 330:302—309,

1971

P, 000DENOUGH DA: Fracture faces of zonula occludentes from "tight" and "leaky" epithelia. J Cell Biol 58:390—400,

for Na-dependent sugar transport. Biochim Biophys Acta 112: 278—291, 1966

43. MARUYAMA T, H05HI T: The effect of D-glucose on the electrical

potential profile across the proximal tubule of newt kidney. Biochim Biophys Acta 282:214—225, 1972 44. ROSE RC, SCHULTZ SG: Studies on the electrical potential profile across rabbit ileum. Effects of sugars and amino acids on transmural and transmucosal electrical potential differences. J Gen Physiol 57:639—663, 1971 45.

amino acid transport. Pflugers Arch 393:179—189, 1982 46. HOSHI T, SAITO Y: Specificity of sodium-dependent electrogenic sugar transport in amphibian kidney proximal tubule. Conrr Nephrol (Karger) 6:52—60, 1977

47. SILVERMAN M, AGANON MA, CHINARD FD: Specificity of monosaccharide transport in dog kidney. Am J Physiol 218:743— 750, 1970

48. SHIMADA T, H0sHI T: A comparative intestinal

27.

TAYLOR RE: Cable theory, in Physical Techniques in Biological Research, edited by WL NASTAK, (vol. VI) Electrophysiological Methods Part B, 1963, pp. 219—262

28. ANAGNOSTOPOULOS T, VELU E: Electrical resistance of cell membranes in Necturus kidney. Pflugers Arch 346:327—337, 1974 29. GUGGINO WB, WINDHAGER EE, BOULPAEI' EL, GIEmscH G:

port of sodium and organic solutes, in Adv Physiol Sci (vol 11), Kidney and Body Fluids, edited by L TAKAC5, Budapest, Perga-

in Amino Acid Transport and Uric Acid Transport, edited by S SILBERNAGL, F LANG, R GREGER, Stuttgart, Georg Thieme Publisher, 1976. pp. 96—104 50. HOSHI T, SuD0 K, SUZUKI T: Characteristics of changes in the intracellular potential associated with transport of neutral, dibasic

and acidic amino acids in Triturus proximal tubule. Biochim Biophys Acta 448:492—504, 1976

51. SAMARZUA I, FROMTER E: Electrophysiological analysis of rat

renal sugar and amino acid transport. IV. Basic amino acids. Pflugers Arch 393:210—214, 1982

52. SAMARZUA I, FROMTER E: Electrophysiological analysis of rat renal sugar and amino acid transport. V. Acidic amino acids. Pflügers Arch 393:215—221, 1982

53. Bss B:

Transport of L-arginine in brush border vesicles derived from rabbit kidney cortex. Arch Biochim Biophys 191:551—560,

mon Press and Akademiai Kiado, 1981, pp. 403—407

1978

31. ANAGNOSTOPOULOS J, TEULON J, EDELMAN A: Conductive prop-

erties of the proximal tubule in Necturus Kidney. J Gen Physiol 75:553—587, 1980

32. CooK DI, FROMTER E: Is the voltage divider ratio a reliable estimate of the resistance ratio of the cell membranes in tubular epithelia? Pflugers Arch 403:388—395, 1985 33. GUNTER-SMITH PJ, GRASSET E, SCHULTZ SG: Sodium-coupled

amino acid and sugar transport by Necturus small intestine. An equivalent electrical circuit analysis of a rheogenic cotransport system. J Membr Biol 66:25—39, 1982 34. PAPPENHEIMER JR. REISS KZ: Contribution of solvent drag through

54. HILDEN SA, SACKTOR B: L-Arginine uptake into renal brush border membrane vesicles. Arch Biochim Biophys 210:289—297, 1981 55. HAMMERMAN MR: Nat-independent L-arginine transport in rabbit

renal

36.

brush border membrane vesicles. Biochim

Biophys Acta

685:71—77, 1982 56.

MCNAMARA PD, REA CT, SEGAL 5: Lysine uptake by rat renal brush border membrane vesicles. Am J Physiol 251:F734—.F742, 1986

57. ULLRICH KJ, RUMRICH G, KLoss S: Sodium dependence of the amino acid transport in the proximal convolution of the rat kidney.

intercellular junctions to absorption of nutrients by the small 35.

HosFil T: Electrophysiological studies on amino acid transport across the luminal membrane of the proximal tubular cells of Triturus kidney,

Cellular and paracellular resistances of the Necturus proximal tubule. J Membr Biol 67:143—154, 1982 30. H05HI T, KAWAHARA K, YOKOYAMA R, SUENAGA K: Change in membrane resistances of renal proximal tubule induced by cotrans-

study of specificity of the

Na/sugar contransport among vertebrates. Comp Bio-

chem Physiol 84A:365—370, 1986 49.

26. CLAUDE 1973

FRöMTER F: Electrophysiological analysis of rat renal sugar and

Pfluigers Arch 351:49—60, 1974

intestine of the rat. J Membr Biol 100:123—136, 1987 MADARA JL, PAPPENHEIMER JR: Structural basis of physiological regulation of paracellular pathway in intestinal epithelia. J Membr

58. LEE SH, PRITCHARD JB: Proton-coupled L-lysine uptake by renal brush border membrane vesicles from mullet. J Membr Biol 75:

Biol 100:149—164, 1987

59. SHALHOUB R, WEBBER W, GLABMAN 5, CANESSA-FISCHER M, KLEIN J, DE HAAS J, Prrrs RF: Extraction of amino acids from and

H05HI T, KIKUTA Y: Effects of organic solute-sodium cotransport on the transmembrane potentials and resistance parameters of the

171—187, 1983

their addition to renal blood plasma. Am J Physiol 204:101—186,

proximal tubule of Triturus kidney, in Electrophysiology of the Nephron, edited by T ANAGNOSTOPOULOS, Paris, INSERM, 1977,

1963 60.

pp. 135—160

cotransport

37. WEBER GH, FROMTER E: Influence of lateral intercellular spaces on

current propagation in tubular epithelia as estimated by multi-cable port of sugars. Fed Proc 21:891—895, 1962

39. CRANE RK: Na-dependent transport in the intestine and other animal tissues. Fed Proc 24:1000—1006, 1965 40. SCHULTZ SG, ZALSKY R: Ion transport in isolated rabbit ileum. II

by renal brush border glutamate transporter. Biochem-

istry 22:5459—5463, 1983

61. HEINZ E, SOMMERFELD DL, KINNE RK: Electrogenicity of sodi-

um/L-glutamate cotransport in rabbit renal brush-border mem-

model. Pflflgers Arch 411:153—159, 1988

38. CRANE RK: Hypothesis for mechanism of intestinal active trans-

NELSON PJ, DEAN GE, ARONSON PS, RUDNICK G: Hydrogen ion

branes: A re-evaluation. Biochim Biophys Acta 937:300—308, 1988 62. WOLFF NA, KINNE R: Taurine transport by rabbit kidney brushborder membranes: Coupling to sodium, chloride, and the membrane potential. J Membr Biol 102:131—139, 1988 63.

GANAPATHY V, LEIBACH FH: Role of pH gradient and membrane

Hoshi: Pitts Memorial Lecture

170

potential in dipeptide transport in intestinal and renal brush border membrane vesicles from the rabbit. Studies with L-carnosine and glycyl-L-proline. J Biol Chem 258:14189—14192, 1983 64. TAKUWA N, SHIMADA T, MATSUMOTO H, HosHI T:Proton-cou-

pled transport of glycylglycine in rabbit renal brush border membrane vesicles. Biochim Biophys Acta 814:186—190, 1985 65. SHIMADA T, HOSHI T: Role of Na/H antiport in intracellular pH regulation by rabbit enterocytes. Biochim Biophys Acta 901:265— 272, 1987

66. KINNE-SAFFRAN E: Transport function of renal cell membranes: ATP-driven transport systems, in Renal Biochemistry, Cells, Membranes, Molecules, edited by RKH KINNE, Amsterdam, Elsevier, 1985, pp. 143—174

67. HEDIGER MA, COADY Mi, IKEDA TS, WRIGHT EM: Expression,

cloning and cDNA sequencing of the Na/glucose co-transporter. Nature 330:379—381, 1987 68. KESSLER M, SEMENZA G:

74. Host-u T, TAKUWA H, ABE M, TA.JIMA A: Hydrogen ion-coupled transport of D-glucose by phlorizin-sensitive sugar carrier in intestinal brush border membranes. Biochim Biophys Acta 861:483—488, 1986

75. SILBERNAGL S: Renal transport of amino acids and oligopeptides, in Renal Transport of Organic Substances, edited by R GREGER, F LANG, S SILBERNAGL, Berlin/Heidelberg/New York, SpringerVerlag, 1981, pp. 93—1 17 76. YOKOYAMA R, H0sHI T: On the relation between velocity parameters and concentration profile of absorbed solutes in a biological

flow-absorption system. (Japanese with English abstract). Jpn J Med Electron Biol Eng 16:373—377, 1978 77. YOKOYAMA R, HosHI T: On the optimality principle in a bio-flow

absorption

relationship in the

system: An analysis of

proximal tubule of the kidney. Pioneers

Biomed Eng (Amsterdam)

(in press)

The small intestinal Na, D-glucose

cotransporter: An asymmetric gated channel (or pore) responsive to 'I'. J Membr Biol 76:27—56, 1983 69. HEINZ E, GRASSL SM: Electric aspects of co- and countertransport, in Electrogenic Transport: Fundamental Principles and Physiological Implications, edited by MP BLAUSTEIN, M LIEBERMAN, New York, Raven Press, 1984, pp. 93—104 70. HosHI T, HIMUAKI M: Na-coupled transport of organic solutes in

animal cells, in Transport and Bioenergetics in Biomembranes, edited by R SATO, Y KAGAWA, Tokyo, Japan Scientific Soc. Press, New York, Plenum Press, 1982, pp. 111—135 71. JAUCH P, PETERSEN OH, LAUGER P: Electrogenic properties of the

sodium-alanine cotransport in pancreatic acinar cells. I. Tight-seal whole cell clamping. J Membr Biol 94:99—115, 1986

72. HosHi T: Electrophysiological aspects of Na-coupled cotransport of organic solutes and Cl in epithelial cell membranes, in Membranes and Membrane Processes, edited by E DRI0LI, M NAKAGAKI, New York, Plenum, 1986, pp. 405—419 73. HILDEN S, SACKTOR B: Potential-dependent D-glucose uptake by

renal brush border membrane vesicles in the absence of Nat Am J Physiol 242:F340—F345, 1982

78. OBERLEITHNER H, GUGGINO W, GIEBISCH G: Mechanism of distal tubular chloride transport in Amphiuma 242:F331—F339, 1982

79. GUGGINO

WB, OBERLEITHNER H,

diluting segment. Am

kidney. Am J Physiol

GIEBIscH

0:

J Physiol 254:F615—F627,

The amphibian

1988

Y, ITol K: Differences in functional properties between the early and the late segments of the distal tubule of amphibian (Triturus) kidney. Jpn J Pharmacol 23:889—896, 1981 81. GREGER R: Ion transport mechanism in thick ascending limb of Henle's loop of mammalian nephron. Physiol Rev 65:760—797, 1985 82. HosHi KURAMOCHI G, SAKAMOTO H: Kinetic properties of active Cl transport across amphibian early distal tubule, in Cou80. HosHI T, SUZUKI

T,

pled Transport in Nephron, edited by T HosHI, Tokyo, Miura Med Res Foundation, 1984, pp. 131—142 83. SAKAM0T0 H, H05HI T: Effects of Ba2 on secretory K movement and electrical properties of the early distal tubule of Triturus kidney. Jpn J Physiol 35:1033—1049, 1985 84. TEULON J, ANAGNOSTOPOULOS T: The electrical profile of the distal tubule in Triturus kidney. Pflagers Arch 395:138—144, 1982 85. SCHWAB A, OBERLEITHNER H: Trans- and paracellular trans-

K

port in diluting segment of frog kidney. Pflugers

1988

Arch 411:268—272,