Acta physiol. scand. 1977. 99. 399-411 From the Institute of Biological Chemistry A, Copenhagen, Denmark

Effect of the Polyene Antibiotic Filipin on the Permeability of the Inward- and the Outward-facing Membranes of the

Isolated Frog Skin (Rana temporaria) BY

ROBERT NIELSEN

Received 16 July 1976

Abstract

NIELSEN, R. Effect of the polyene antibiotic filipin on the permeability of the inward- and the outward-facing membranes of the isolated frog skin (Rana temporaria). Acta physiol. scand. 1977. 99. 399411.

The effect of the polyene antibiotic filipin on the electrical properties and the passive permeability of the cell membranes was investigated. The addition of filipin to the outside bathing solution has the following effects: 1. it results in a drastic reduction in the transepithelial resistance and potential, 2. it causes a 10-20 times increase in the passive transepithelial chloride, sodium and sucrose flux, 3. it results in the forma- tion of an amiloride insensitive sodium pathway in the outward facing membrane, 4. it results in an active outward transport of potassium, 5 . it results in a highly significant swelling of all the cells in the epithelium. The addition of filipin to the inside bathing solution has the following effects 1. it results in an activation of the active sodium transport, 2. it causes a slight increase in the passive transepithelial chloride and so- dium permeabilities but has no effect on the sucrose permeability, 3. it has no effect on the amiloride in- hibition of the short-circuit current, 4. it has no effect on the volume of the cells in the epithelium. It is suggested that the addition of filipin to the outside bathing solution increases the direct sodium flow from cell to cell in neighbour layers. Furthermore these experiments indicate that the outward facing membrane of the isolated frog skin has a high cholesterol content as compared with the cholesterol content of the inward facing membrane.

The polyene antibiotics are characterized by a large polyhydroxy lactone ring of 23-43 atoms with 4 to 7 conjugated double bonds in the ring. The polyene antibiotics filipin and amphotericin B mediate changes in the permeability of bilayer membranes, liposomes, erythrocytes, fungi and intestinal mucosal cells from chicks (for ref. see Norman et al. 1972).

In sodium transporting epithelia, such as the frog skin and the toad bladder, the addition of amphotericin B to the outside (the mucosal side) bathing solution of the epithelia re- sulted in an increase in the passive chloride and urea permeabilities, whereas the effect on the active sodium transport was variable (Lichtenstein and Leaf 1965, Finn 1968, Nielsen 1971). The addition of amphotericin B to the inside of the frog skin and the serosal side of the toad bladder had no effect on the active sodium transport (Lichtenstein and Leaf 1965,

399

400 ROBERT NIELSEN

Finn 1968, Nielsen 1971). On the basis of these and other experiments Lichtenstein and Leaf ( 1965) concluded that amphotericin B increased the permeability of a dense diffusion barrier near the mucosal surfac? to small solutes, whereas Finn (1968) concluded that amphotericin B increased the passive movement of sodium and small solutes across the membrane facing the medium to which the drug was added.

In a previous paper I have suggested a model to explain the changes in the sodium transport during the aldosterone induced moult (Nielsen 1972, 1973). This model was partially based on the assumption that addition of filipin o r amphotericin B to the outside bathing solution rcsulted in an increase in the passive transepithelial sodium flux.

In the present studies it is shown that addition of filipin to the outside bathing solution results in a 10-20 times increase in the passive transepithelial sodium, chloride and sucrose fluxes. Furthermore the addition of filipin to the inside bathing solution results in a n in- crease in the active sodium transport but has nearly n o effect on the passive transepithelial sodium, chloride and sucrose fluxes.

Methods

The cvperimcnts wcre performed on male and female frogs (Rana temporaria). The frogs were kept par- tially immersed in tap water at about 4 C . The skins were dissected from pithed animals and divided into t A O laterally symmetrical pieces. The skins were mounted i n perspex chambers and bathcd in stirred Ringer's solutions (Na-

The short-circuit experiments were performed according to the method of Ussing and Zerahn (1951), using an automatic voltage clamp apparatus programmed to disconnect the short-circuit current every fibe minutes, thus allowing the potmtial to be measured for 60 s. The fluxes are in some experiments expressed in t x m s of permeability coefficients, calculated by adding the isotope to the medium bathing o:i? side and measuring its rate of appearance on the other side (Andersen and Ussing 1957).

115.0, K+ ~~ 2.5, Ca'--~ 1.0, HCOT --2.4. CI-- ~ 117.1, mM, p H - 8 . 2 ) .

P ~ (a,V, ~ a,V,)/A as (t, - t,)

ill a n d a? designat. the radioactivity originating from side 2 in one ml of solution from side 1 betwecn times t, and t,, respectively, V, and V, th- corresponding volumes of the solution on side 1, A the area of the membrane and as tlie mean activity in one ml of solution from side 2 during the period from t, to t2. The filipin Lot u-5956 was a gift from the Upjohn Company. Kalaniazoo, Michigan, U.S.A.

Results

The effect of .filipin on the shwt-circuit current (SCC) and transepithelial potential ( P D )

Addition of filipin to the outside bathing solution. Fig. 1 A shows the effect of adding 5 / i M filipin to the medium bathing the outside of the isolated frog skin, Fig. 1 B shows the control skin half. 10-40 min after the addition of 5 p M filipin the PD and resistance started to decrease. Higher conccntrations of filipin (50 I'M) caused the decrease in PD and resistance to start 2-3 min after the addition of the filipin. 0.5 pM had no effect on the PD and the resistance. 5 / r M filipin caused a 40-750,; reduction in the PD. 50 pM filipin caused a 75-90% reduction in the PD. The effect of filipin on the SCC was variable. In about half of the experiments filipin had no effect on the SCC o r caused 10-50% reduction in the SCC, in the remaining experiments filipin caused 10-1001; activation of the SCC. This activation of the SCC lasted 1-2 h.

EFFECTS OF FILIPIN ON FROG SKIN

3

50--10

2

25-- 1

401

5

I 0-0 1

l 5 I 5 OLD I 2 3

H",',.

Fig. 1 A and B. Addition of filipin to the outside bathing solution. Effect of filipin on the short-circuit current and potential across the frog skin. A. At the arrow filipin was added to the outside to give a con- centration of 5 pM. B. Control skin half. -, short-circuit current (,uA/cm2); 0-0, potential (mV); .-a, resistance (Kohm x cm2).

Kinsky (1963) has shown that 3 pM filipin caused complete hemolysis of rat erythrocytes, below 1 pM no hemolysis occurred. Van Zutphen et al. (1971) have shown that 0.4 pM filipin caused about 95 'j6 reduction in the resistance of lecithincholesterol (1/1) bilayer mem- branes. In the presence of 0.4 pM filipin the lecithin cholesterol bilayer membrane was stable, in the presence of 4 pM filipin the survival time of the membrane was about 6 min, and in the presence of 40 pM the survival time of the membrane was less than 1 min (Van Zutphen et al. 1971). Thus the concentration of filipin that induced a reduction in the PD and the resistance across the isolated frog skin, hemolysis of rat erythrocytes, and a decrease in the survival time of bilayer membranes was the same. Therefore the decrease in PD and resistance in the isolated frog skin after the addition of filipin to the outside may be due to a partial disintegration of the outward-facing membrane.

Addition of filipin to the inside bathing solution. Addition of 5 pM filipin to the inside had no effect on the PD and the SCC. About 5 rnin after the addition of 50 pM filipin the SCC and to a lesser degree the PD increase (Fig. 2 A). The increase in SCC was 10-100% of the control level. In half the expts. the current response to filipin addition was triphasic, the first maximum in SCC occurring 2&45 min after addition of filipin; thereafter the SCC started to decline to about the control level. After this decrease the SCC increased again and reached its second maximum after about 2 h of incubation.

The effect of filipin on sodium flux

To ascertain whether the increase in the SCC after addition of filipin to the inside of the frog skin was due to an increase in the active sodium flux, a series of double labelling expts. was performed. The net sodium flux and the SCC were identical both in the control period and in the two periods after the addition of filipin (Table I). Thus, the addition of filipin to

26 - 715874

402

A k n c m 2

5

L

- 3

2

1

0

ROBERT NIELSEN

Hours

PA/crn;

LO

30

20

10

0

mV

100

2 5

O A I 2 3 I,

Hours

Fig. 2 A a n 3 I). Addition of filipin to the insidc bathing solution. Effect of filipin on the short-circuit current ariJ potentizl zcross thc frog skin . A . At the arrow filipin was added to the inside to give a con- centrdiion of 5 0 !IM. B. Control skin half. --, short-circuit current (pA/cm'); 0--3, potential (mV); Q ---@. rcsistitiic1' (Kohm c:n2j.

the inside bathing solution of the frog skin rcsults in an activation of the active sodium transport.

Effect on passioe fluxes

To evaluate the effect of filipin on the permeability of the frogskin to substances, the transport of which can be accounted for by passive forces, transepithelial sucrose and chloride in-

EFFECTS OF FILIPIN ON FROG SKIN 403

TABLE I. The net sodium flux and short-circuit current across the frog skin before and after the addition of filipin (50 pM) to the inside of the frog skin.

Period Influx Outflux Net flux SCC

peqlcm21h Control 1.16 0.05 1.11 1.12 0.603'p>0.50 Filipin 1.h. 1.33 0.06 1.27 1.30 0.60>p>-0.50 Filipin 2.h. 1.26 0.04 1.22 1.19 0.50 3' p 3'0.40

~~

The influx was measured by means of Na-22 and the outflux simultaneously with Na-24. The SCC was measured and recorded automatically. The mean results from 7 expts. are presented. Each period was 1 h.

and outfluxes were determined. The sodium outflux (which is passiv) was also measured. Table I1 shows that the addition of filipin to the outside bathing medium, (the skins were mounted in Ringer's solution containing 1 mM sucrose) results in large and sustained increases in the influx and outflux of sucrose, whereas the addition of filipin to the inside had no effect on these fluxes. The addition of filipin to the outside bathing medium results in a large and sustained increase in the chloride in- and outflux, and in the sodium outflux (Table 111). Furthermore there was no significant difference between the chloride in- and outflux. The addition of filipin to the inside bathing medium had only a small effect on the chloride in- and outflux and on the sodium outflux (Table IV). Fig. 3 shows a series of experiments in which the sodium outflux is plotted against the chloride outflux. The out- fluxes of sodium and chloride were measured simultaneously on the same skin after the addition of filipin to the outside bathing medium. The slope of the line drawn in Fig. 3 is equal to the ratio of the free solution mobilities of sodium and chloride i.e. 0.66. The experi- mental points fit the theoretical line. This indicates that the addition of filipin to the outside bathing medium results in the formation of a nonspecific transepithelial pathway.

Previous data (Nielsen 1971) have shown that the addition of 50 pM amphotericin B to the outside bathing solution results in a 5 times increase in the sodium outflux and a 5 times increase in in- and outflux of chloride and urea. The increases in the chloride and urea in- and outfluxes were essentially the same in both directions. The addition of amphotericin

TABLE 11. Effect of 50pM filipin on the permeability of the frog skin to sucrose.

Period Control Filipin i S.E. 1 S . E . A-B

B-C C-D D-E

Fiiipin on cm x sec-' x lo-' outside 0.87t0.24 7.76k0.54 10.6k 0.6 13.71 1.7 inside 0.61 i0.07 0.5710.1 1 0.59k0.10 0.70i0.09

In each expt. following 20 min of equilibration for the isotope the permeability across the tissue was deter- mined for one, 1 h control period (period A-B). Filipin was then added to the outside or the inside bathing medium to give a concentration of 50 pM. After the addition of filipin solution, the permeability was determined for 3 further 1 h periods (periods B-C, C-D, and D-E). Both the sucrose in- and outflux were measured in these expts. The results were essentially the same for fluxes in either directions. All values are the meansiS.E. of 6 expts.

404 ROBERT NIELSEN

TAKE 111. Effect of filipin on the permeability of the frog skin to chloride and sodium. 50 p M filipin in the outside bathing medium.

Period Number of Control Filipin expts. 1- S.E. - __ S.E.

A-B B-C C-D

cm sec-li lo-' Chloride outflux 8 2.16k 0.55 28.1 & 4.9 43.1 f 5 . 3 Chloride influx 9 3.0620.46 21.3k5.9 32.9k9.9 Sodium outflux 5 1 . 0 9 ~ 0.41 19.6k5.5 24.6k 6.4

In each expt. following 20 min of equilibration for the isotope the permeability across the tissue was deter- mined for one, I h period (period A-B). Filipin was then added to the outside bathing medium. After the addition of the filipin solution, the permeability was determined for 2 further 1 h periods (periods B-C and C D). All values are the means: S.E.

B to the inside bathing solution has no effect on the passive chloride and urea flux (Table V). For both substances in- and outflux were measured. Thus the addition of amphotericin B and filipin to the outside bathing solution resulted in a significant increase in the passive transepithelial fluxes, for the substances tested, whereas the addition of amphotericin B and filipin to the inside bathing solution had no or very small effects on these fluxes.

Actioit of amiloride

Precious studies have shown that addition of amiloride to the outside bathing solution of the frog skin and the toad bladder inhibit the active transepithelial sodium transport (Salako and Smith 1970, Nielsen and Tomlinson 1970, Bentley 1968). This inhibition is believed to be caused by an interaction of amiloride with the specific sodium pathway (the sodium channels) in the outward facing membrane (Cuthbert and Shum 1974).

Fig. 4 shows an experiment in which the SCC was blocked by amiloride. The addition of filipin to the outside bathing solution caused full recovery of the SCC. In 10 expts. the addi- tion of 50 pM filipin to the outside bathing solution after the addition of amiloride resulted in 20-100°,, recovery of the SCC. Thus addition of filipin to the outside bathing solution

TABLE IV. Effect of filipin on the permeability of the frog skin to chloride and sodium. 50pM filipin in the inside bathing medium.

Period Number of Control Filipin expts. 5 S.E. I S.E.

A-B B-C C-D

cm sec-' Chloride outflux 7 2.26k 0.48 3.04k0.59 3.492 0.75 Chloride influx 7 1 . 8 4 i 0.34 2.38 0.43 2.51 k0 .44 Sodium outflux 5 I . I9 2 0.22 1.66-t 0.16 1.71 & 0.66

In each expt. following 20 min of equilibration for the isotope the permeability across the tissue was deter- mined for one, 1 h period (period A-8). Filipin was then added to the inside bathing medium. After the addition of the filipin solution, the permeability was determined for 2 further 1 h periods (periods B-C and C-D). All values are the mzaiis2S.E.

EFFECTS OF FlLlPIN ON FROG SKIN 405

Fig. 3. The sodium outflux is plotted against the chloride out- flux. The outfluxes of sodium and chloride were measured simul- taneously on the same skin after the addition of filipin (50 pM) to the outside bathing medium.

results in the formation of a pathway in the outward facing membrane that is insensitive to amiloride.

Fig. 5, which is representative for 8 expts., shows that the addition of filipin to the inside bathing solution has no effect on the SCC when this was blocked by amiloride. After the outside bathing solutions were replaced by solutions without amiloride, the control skin half showed the normal filipin induced changes in the SCC. Thus the effect of filipin added to the inside bathing solution is not due to a formation of an amiloride insensitive sodium pathway, but is the result of an increase in the sodium flux via the amiloride sensitive sodium pathway.

Effect on cell volume

Although addition of filipin to the inside had a very small effect on the transepithelial per- meability, it remained still an open question whether filipin caused a large increase in the permeability of the inward facing cell membranes. If filipin caused a large increase in the non-specific permeability of the inward facing cell membranes, the cells would swell. To

TABLE V. Effect of amphotericin B on the permeability of the frog skin to chloride and urea. 50pM ampho- tericin B in the inside bathing solution.

Period Number of Control Amphotericin B expts. ? S.E. & S.E.

A-B B-C C-D

Urea Chloride

cm x sec-' x lo-' 5 5.38k0.57 5.391 0.60 5.26k0.62 6 5.001- 0.61 4.57+ 0.75 4.93+ 1.58

In each expt. following 20 min equilibration for the isotopes the permeability across the skin was deter- mined for one, 1 h control period (period A-B). Amphotericin B was then added to the inside bathing medium. After the addition of the amphotericin B solution, the permeability was determined for 2 further 1 h periods (period B-C, C-D). All values are the means? S.E.

406 ROBERT NIELSEN

I / 1.7pM AMlLORlDE

I , 1 0 1 2 3

Hours

Fig. 3. Effect of filipin (outside) on the SCC of a frog skin where the SCC was blocked by aniiloride. At the arrows amiloride was added to the outside bathing solution of the skin halves to give a concentration of 1.7!tM. At the other arrow filipin (50,1(M) was added to the outside of the skin half represented by the \slid line.

i */cm2

80

40 - _ _ _

20 -

I I I I 2 O L 4 6 tl

Hours

Fig. 5. Effect of filipin (inside) on the SCC of a frog skin where the SCC was blocked by amiloride. A1 the arrows amiloride was added to the outside bathing solutions of the skin halves to give a concentration of 1.3 ItM. At the dotted arrow 50 p M filipin was added to the inside bathing solution of the skinhalf rep- resented by the dotted line. After 3 h of incubation the outside bathing solutions were replaced by fresh Ringer's solution and at the arrow (4 h) 50 / iM filipin was added the inside bathing solution of the skin half represented by the solid line.

EFFECTS OF FILIPIN ON FROG SKIN 407

Fig. 6. Light micrographs of frog skin epithelium incubated with 50 ,uM filipin in the inside- or the outside bathing solution, 1 pm sections stained with toluidine blue. The bar on Fig. 6 a corresponds to 10 pm. 6 a. The skin is incubated in 1 h with filipin in the outside bathing solution a marked swelling of the cells in the 1. RCL is observed. 6 b. The skin is incubated in 3 h with filipin in the outside bathing solution, a marked swelling of the outermost 4-5 cell layers is observed, and the interspace system is closed by the swollen cells. 6 c. Control skin. 6 d. The skin is incubated in 3 h with filipin in the inside bathing solution. The cells in the lower cell layers did not show any tendency to swell. Concerning the effect of filipin on the cells in the 1. RCL see text.

test this possibility expts. were performed in which the skins were incubated with filipin on the inside or on the outside from 10 min to 3 h. After the incubation, the skins were fixed by adding Ringer’s solution with OsO, to both chamber halves, to give a concentration of 1 % OsO, in the solutions bathing the skin. Fig. 6 (a, b) (representative for 7 expts.) shows the effect on cell volume of adding 50 pM filipin to the outside bathing solution. After 60 min incubation the volume of all the cells increases (Fig. 6 a) in the outermost layer of the stratum grandasurn (the first reacting cell layer, 1. RCL). After 3 h incubation the cells of the underlaying 1 4 layers are swollen to such an extent that the interspace system between them appears closed (Fig. 6 b). Saladino, Bentely and Trump (1969) have shown that the addition of amphotericin B to the mucosal side of the toad bladder causes a swelling of the cells-in the epithelium, and Swelto, Perrini and Lippe (1975) have shown the addition of amphotericin B to the outside of the frog skin results in a swelling of the cells in the 1. RCL. Even after 3 h of incubation with 50 pM filipin in the inside bathing solution the cells in the lower cell layers did not swell (Fig. 6 D, representative for 10 expts.). The effect of filipin (inside) on the cells in the 1. RCL was variable. In seven skins it had no effect, in two skins

408 ROBERT NIELSEN

it caused slight swelling-and in one skin a slight shrinkage. Hence addition of filipin to the outside results in an increase in the permeability of the outward facing cell membranes, whereas the addition of filipin to the inside had no or a very small effect on the sodium permeability of the inward facing cell membranes.

Discussion

From the data presented above it appears that the addition of filipin to the outside bathing solution resulted in an increase in the passive transepithelial sodium, chloride, and sucrose fluxes and in a progressiv swelling of the cells in the different cell layers. The addition of filipin to the inside bathing solution however had no effect on the passive transepithelial fluxes and no effect on the volume of the cells. The discussion will be separated in two parts. A, the effect of filipin on the transepithelial fluxes and B, the effect of filipin on the progres- sive swelling of the cells in the different cell layers.

According to the two membrane hypothesis (Koefoed-Johnsen and Ussing 1958) the frog skin can be treated as composed of an “outward-facing membrane” which is selectively permeable to sodium ions but impermeable to potassium ions and permeable in a non- selective way to small anions like chloride. The “inward-facing membrane” is permeable to potassium and small anions, but impermeable to free sodium ions. The outward-facing membrane is located just beneath the cornified layer (Ussing and Windhager 1964, Farquhar and Palade 1966). The outward facing membrane consists of the outward-facing cell mem- branes of the outermost layer of the s. granulosum (1. RCL) kept together by tight seals. The inward-facing membrane is identical with the cell membranes limiting the extracellular space throughout the epithelial layer (Fig. 7 A). Thus a solute can permeate across the skin via two types of “pathways”, via an extracellular route or via an intracellular route. A solute that permeates via the intracellular route has to cross both an inward- and outward-facing cell membrane (Fig. 7 A). A component that only increases the sodium permeability of the outward-facing cell membranes should therefore have no effect on the passive sodium out- flux (the inward-facing membranes is impermeable to free sodium), whereas a component which increases the sodium permeability of the inward-facingImembrane should increase the passive sodium outflux.

Addition of filipin to the outside results in an increase in the amiloride insensitive sodium transport (Fig. 4), thus filipin increases the sodium permeability of the outward facing cell membrane. Furthermore it has been shown by Nielsen (1971, 1972) that filipin and ampho- tericin B increase the potassium permeability of the outward facing cell membrane, too. An increase in the sodium and potassium permeability of the outward facing membrane would result in an increase in the cellular sodium concentration and a decrease in the cellular potassium concentration. These changes in the cellular compositions should result in an increased sodium transport. Such an increase was only observed in half the experiments. But if the addition of filipin to the outside bathing solution makes the outward facing mem- brane so permeable to sodium that the sodium flux into the cells exceeds the capacity of the sodium pump, the cells would undergo colloid-osmotic swelling and eventually burst. The

EFFECTS OF FILIPIN ON FROG SKIN 409

A B C D

Fig. 7. Diagram of the 3 outermost layers of the isolated frog skin. The top layer (the hatched layer) is the s. corneum. The underlaying layer is the s. granulosum (I. KCL). Thc 1. RCL ( I . reacting cell layer) is selectiv permeable to sodium (indicated by the dots). 7 A. OFM, outward facing membrane. I, intracellular route. 7 B. a and b, disintegrations of the outward (a) and (h) the inward facing cell membranes. 7 C. G , gap junction formed as a result of the filipin induced swelling of the 1. RCL. For further information see text. 7 D. Filipin induced changes in the 1. RCL could result in formation of a new outward facing cell membrane (OFM,,) in the underlaying cell layers. For furthcr information see text.

variable effect of filipin on the SCC could readily be understood if filipin in some skins results in a fast and extensive cytolysis of the cells. The data of Fig. 6 A show that addition of filipin to the outside solution results in a swelling of the cells.

This swelling of the cells would most likely result in a partial disintegration of both the inward and the outward facing cell membranes (Fig. 7 B). A partial disintegration of the inward and outward facing cell membranes would result in a nonspecific increase in the passive transepithelial permeability, which could explain the increase in the passive trans- epithelial fluxes of sucrose, chloride and sodium (Tables 11, 111 and Fig. 3).

Since the addition of filipin to the inside bathing solution has no effect on the sodium outflux, filipin (according to the two membrane hypothesis explained above) has no effect on the sodium permeability of the inward facing cell membrane, and therefore it should have no effect on thz cell volume and the passive transepithelial permeability. That this interpreta- tion is correct appears from the data of Tables I t , IV and Fig. 6 D. The addition of filipin to the inside results, however, in an increase in the active sodium transport (Fig. 2, Table 11). This is probably caused by an indirect effect of filipin on the potassium permeability of the inward facing membrane, mediated by calcium entry (Nielsen 1976).

Many studies show that the presence of sterols in the membrane is a requirement for poly- ene antibiotic sensitivity (for references see Norman et al. 1972). If as suggested by Kruijff and Demel (1974) the filipin induced permeability changes are a result of filipin forming large filipin cholesterol complexes in the mcmbrane, the outward-facing membrane must be assumed to have a high cholesterol content, bccause filipin per se only increases the non specific permeability of the outward-facing nicmbrane; the cholesterol content of the in- ward-facing membrane must be low.

After 1 h of incubation with filipin (outside) the cells of the 1. RCL are swollen and have a necrotic appearance (Fig. 6 A); none the less most of the skins are still able to generate a high SCC, whereas the skins are unable to generate a SCC when nearly all the cells in the

410 ROBERT NIELSEN

epithelium are swollen. Therefore, presumably, after 1 h of incubation the I . RCL consists of cells which are unable to produce a SCC, and the cells in the underlaying cell-layers generate the SCC. Under these circumstances sodium can diffuse easily from or through the cells in the 1. RCL to the cells in the underlaying layers. However, Voiite and Ussing (1968) have shown that only the 1. RCL swells during short-circuiting. This result, and the kinetic analysis of sodium and lithium movements in the isolated frog skin by Morel and Leblanc (1973) show that the diffusion from cells in one layer to those in the adjacent ones is very slow. This implies that in normal frog skin the cells of the 1. RCL are responsible for the major part of the active transepithelial sodium transport. That is supported by expts. of VoClte and Ussing (1970) which show that expansion of the interspace system have n o effect on the SCC (in these expts. the interspaces (apart from the tight seals) were so expanded that the cells only were in contact via the desmosomes). Since filipin per se only increases the non specific permeability of the outward-facing membrane of the 1 . RCL, this effect cannot explain the increase in the direct sodium flow from cell to cell in neighbouring layers. The following two models ( A and B) can explain the increase in the direct passing of sodium from one cell layer to the next. A, when the cells in the 1 . RCL swell, they come in contact with the cells of the underlaying cell layer resulting in the formation of new gap junct ions (Fig. 7 C) or B, the filipin-induced changes in the 1 . RCL could result in the release of a signal which caused a change in the outward-facing membrane of the underlaying cell layer, so that this cell layer becomes sensitive to filipin Fig. 7 D (e .g . these cells might form a new outward-facing sodium selective membrane). This hypothesis can explain how an epithelium with all the cells in the 1. RCL having a marked necrotic appearance still can transport sodium and it can explain the observed progressive swelling of the underlaying cell layers.

Experiments are in progress t o investigate whether the filipin induced necrosis of a cell layer results in the formation of gap junctions o r results in the formation of a new outward facing cell membrane of the cells facing the necrotic cells.

References

ANDEKSEN, B. and H. H . USSIWG, Solvent drag on non electrolytes during osmotic flow through isolated

BENTLEY, P. J., Amiloride: a potent inhibitor of sodium transport across the toad bladder. J. Physiol.

CUTHBERT, A . W. and W. K . SHUM. Binding of amiloride to sodium channels in frog skin. Mol. Pharmacol.

De KRIJIIFF, B. and R. A. DEMEL. Polyene antibiotic-sterol interactions in membranes of acholeplasma

F ~ R Q U H A R , M. G. and G . E. PALADE, Adenosine triphosphatase localisation i n amphibian epidemis.

FINK. A . L.. Separate effects of sodium and vasopressin on the sodium pump in toad bladder. Amer. J .

K I ~ S K Y . S. C., Comparative responses of mammalian erythrocytes and microbial protoplasts to polyene

toad skin and its response to antidiuretic hormone. Acta physiol. scand. 1957. 39. 228-229.

(Lond.) 1968. /95. 317-330.

1974. 10. 880-891.

laidlawii cells and lecithin liposomes. Biochini. biuplt.r.s. Acra (Amst.) 1974. 339. 57-70.

J . cell. Bid. 1966. 30. 359-379.

Physiol. 1965. 225 . 849-856.

antihiotica and vitamin A. Arch. Bioclwni. 1963. 102. 180-188.

Loewenstein (1973) has shown that junctional passageways (gap junctions) form rapidly where cells, that can make them, come into contact.

EFFECTS OF FILIPIN ON FROG SKIN 41 1

KOEFOED-JOHNSEN, V. and H. H. USSING, The nature of the frog skin potential. Acta physiol. scand. 1958.

LICHTENSTEIN, N. S. and A. LEAF, Effect of amphotericin B on the permeability of the toad bladder. j . elin. Invest. 1965. 44. 1328-1342.

LOEWENSTEIN, W. R., Cell coupling. In Transport Mechanisms in Epithelia. Alfred Benson Symposium v , eds. H. H. Ussing and N. A. Thorn. Munksgaard, Copenhagen 1973. 20-26.

MOREL, F. and G. LEKLANC, Kinetics of sodium and lithium accumulation in isolated frogskin epithelium. In Trnnsport Mechanisms in Epithelia. Alfred Benson Symposium V, eds. H. H. Ussing and N. A. Thorn. Munksgaard, Copenhagen 1973. 73-82.

NIELSEN, R., Effect of amphotericin B on the frog skin in vitro. Evidence for outward active potassium transport across the epithelium. Acra physiul. scund. 1971. 83. 106-114.

NIELSEN, R., The effect of polyene antibiotics on the aldosterone induced changes in sodium transport across isolated frog skin. J . Ster. Biuchem. 1972. 3. 121-128.

NIELSEN, R., Effects of aldosterone on frog skin. In Transport Mechanisms in Epithelia. Alfred Benson Symposium V, eds. H. H. Ussing and N. A. Thorn. Munksgaard, Copenhagen 1973. 214-224.

NIELSEN, R., Does the polyene antibiotic filipin act as a Ca ionophore, when added to the inside of the isolated frog skin? Actu physiol. scand. 1976. Suppl. 440. 72.

NIELSEN, R. and R. W. S. TOMLINSON, The effect of amiloride on sodium transport in the normal and moulting frog skin. Acta physiol. scand. 1970. 79. 238-243.

NORMAN, A. W., R. A. DEMEL, B. DE KRUIYFF, W. S. M. GEURTS VAN KESSEL and L. L. M. VAN DEENEN, Studies on the biological properties of polyene antibiotics: Comparison of other polyens with filipin in their ability to interact specifically with sterol. Biochim. biuphys. Acta (Amst.) 1972. 290. 1-14.

RAWLINS, F., L. MATEU, F. FRACACHAN and C. WHITTEMBURY, Isolated toad skin epithelium: Transport characteristics. Pjlugers Arch. ges. Physiol 1970. 316. 64-80.

SALADINO, A. J., P. J. BENTLEY and B. F. TRUMP, Ion movements in cell injury. Amer. J . Puthol. 1969.54. 42 1-466.

SALAKO, L. A. and A. J. SMITH, Changes in sodium pool and kinetics of sodium transport in frog skin produced by amiloride. Brit. J. Pharmacol. 1970. 39. 99-109.

SVELTO, M., M. C. R. PERRINI and C. LIPPE, Anomalous effects of amphotericin B on the non-electrolyte fluxes through the skin of Rana Esculenta. Gen. Pharmacol. 1975. 6. 105-108.

USSING, H. H. and K. ZERAHN, Active transport of sodium as the source of electric current in the short- circuited isolated frog skin. Acru pliysiol. scund. 1951. 23. 110-127.

USSINC, H. H. and E. WINDHAGER, Nature of the shunt path and active sodium transport path through frog skin epithelium. Acta physiol. scand. 1964. 61. 484-504.

VAN ZUTPHEN, H., R . N. DEMEL, A. W. NORMAN and L. L. M. VAN DEENEN, The action of polyene anti- biotics on lipid bilayer membranes in the presence of several cations and anions. Biochim. biophys. Acta (Amst.) 1971. 241. 310-330.

V O ~ T E , C. L. and H. H. USSING, Some morphological aspects of active sodium transport. J. cell Biol. 1968. 36. 625-638.

VO~ITE, C. L. and H. H. USSING, Quantitative relation between hydrostatic pressure gradient, extracellular volume and active sodium transport in the epithelium of the frog skin (R. Temporaria). Exp. Cell Res. 1970. 62. 375-383.

42. 298-308.