Changes in cation contents of stria vascularis with ouabain and potassium-free perfusion

Hearing Research,4 (1981) 149-160 o Elsevier/North-Holland Biomedical

149 Press

CHANGES IN CATION CONTENTS POTASSIUM-FREE

OF STRIA VASCULARIS

WITH OUABAIN AND

PERFUSION*

DANIEL C. MARCUS, NANCY Y. MARCUS and RUEDIGER THALMANN Washington University Medical School, Dept. of Otokzryngology. 517 So. Euclid Ave., St. Louis, MO 631 IO, U.S.A. (Received 8 October 1980; accepted

12 November 1980)

Perfusion of the perilymphatic space of guinea pig cochleae with K-free medium leads to a gradual decline of the endocochlear potential (BP) over SO-50 min to a negative value (mean: - 12 mv). The input resistance of scala media does not decrease during this time. The ATP and K content of the stria vascularis are reduced by similar amounts (26 and 34%, respectively) during this period. Perfusion of 1 mM ouabain produces a different pattern of response: strial ATP remains normal while strlal K content is strongly reduced (by 77%). Strlal Na rises in a complementary way to the K loss. These results demonstrate that a reduction of the K concentration of the perilymph leads to an inhibition of the generator of the positive component of the BP rather than to a general increase of cochlear duct membrane conductance. In addition, they suggest, in concert with other considerations (such as the slower rate of decline of the EP during K-free vascular perfusion (Wada, J., Kambayashi, J., Marcus, D.C. and Thalmann, R (1979): Arch. Otorhinolaryngol. 225, 79-81) ), that the mode of action may be different from that of ouabain. In spite of the lack of teleological support, we offer the hypothesis that the strial generator of the EP may PrimarilYutilize K from perilymph and that vascular K may not have access to the generator. Key words: stria vascularis; ion content; potassium pump; helium-glow photometer.

INTRODUCTION

We have previously reported that vascular perfusion of the surviving ear with K-free artificial blood produces a reduction of the endocochlear potential (EP), but only after a surprisingly long delay of more than 20 min [20] . In addition, the ensuing decline rate is substantially. slower than that observed after removal of K from perilymph [6, 91. These findings are of interest in light of the dependence of inner ear function upon rheogenic K-transport by the stria vascularis into the cochlear duct. These results seemed at odds with two other observations: (1) The vasculature is apparently in more intimate contact with the cells of the stria vascularis than is the per& lymph. One might presuppose, therefore, that the composition of the extracellular fluid of the stria is likely to be controlled more directly by the blood than by the perilymph.

* This work was presented in part at the meeting of the Acoustical Society of America, April 1979.

150

K-free blood would thus be expected to produce the greater effect, contrary to observation. (2) Ouabain abolishes the positive EP more rapidly via the vascular route than via the perilymph [21] . Since ouabain’s primary action is inhibition of Na/K-ATPase by competition with K for specific membrane sites, one would expect the removal of K also to act more strongly from the vasculature. The present experiments were conducted on the premise that the reduction of the EP by ouabain and by K-free perfusion would differentially alter the cation content of the stria. The perilymphatic, rather than the vascular, route was chosen for perfusion since the effect of K removal here is more dramatic. Although the effects of ouabam on cochlear function have been reported in some detail [ 1, 7, 81, several questions remain concerning the effects of K-free perfusion. In particular, we include here an investigation of: (1) the time course and the amount of reduction of the EP; (2) the relationship between the decline of the EP and the resistance of the cochlear duct; (3) to what extent the decline reflects changes in the K content of the stria vascularis; (4) whether the change of the EP is due to interference with energy generation, as reflected by the ATP content of the stria vascularis. METHODS

Data were obtained from guinea pigs (200-250 g) which were anesthetized with sodium pentobarbital (35 mg/kg, i.p.). The cochleae were exposed by the usual ventrolateral approach. The EP was recorded by means of a glass micropipet electrode inserted into Scala media of the second turn via the spiral ligament and stria vascularis. The input impedance of Scala media was measured by two pairs of electrodes, one pair serving for the injection of current, and the other for recording the potential. Each pair consisted of a glass pipet (filled with 150 mM KCl, tip diameters of 2-5 pm, and electrical connection made with Ag/AgCl) and a Ag/AgCl/lSO mM NaCl reference electrode on the neck muscles. The glass pipets were inserted adjacent to each other. EP measurements were corrected for liquid junction potentials and pressure-induced changes. The waveform of the injected current for impedance measurements was a bipolar square wave of f 1.3 PA at 0.2 Hz. Both perilymphatic scalae were perfused at 5 pl/min from fust turn Scala tympani to first turn Scala vestibuli. The glass cannula at the inlet was sealed to the hole with partiallydried laquer. The control medium consisted of (in mM): NaCl 125, KC1 4, NaHC03 25, MgCls 1.l, NaH2P04 0.8, glucose 3, urea 2, CaCls 1.3; OZ/COZ 95/5%; pH 7.4. The Kfree medium was identical except that additional NaCl was substituted for KCl. Tissues for chemical analyses were prepared as previously described [12]. Briefly, the cochlea was frozen in situ with Freon at - 150°C, separated from the head and lyophilized in toto at -4O’C. The lyophilized stria and spiral ligament from the proximal region of the second turn were dissected and analyzed for the desired constituents. The control cochleae were frozen after 60 min of perfusion and the K-free and ouabain-treated cochleae when the EP was maximally reduced (about 30-50 min following initiation of perfusion). ATP was determined by an enzymatic cycling method, routinely used in this labora-

151

tory [12]. The ion content of the tissues was determined by helium-glow photometry. For this purpose the lyophilized samples (0.2-0.5 pg) were ashed at 5OO’C for 4 h on dented platinum or nickel foil (microcrucibles). The ash was then dissolved in 50 nl of 100 mM HNOa. Subsequently the samples were diluted tenfold. Because of difficulties using the usual CsNOs /NH4 HP04 diluent [ 191 with tissue samples (l-4 mM Na or K), one of two procedures was then followed: (1) Samples were subdivided for separate Na and K measurements. The diluent for K-readings was 30 mM NaCl and that for Nareadings was 30 mM KCl; thus Na and K were determined in separate ‘runs’ of the photometer. (2) Dissolved tissue samples and standards were diluted in 30 mM IXl. In this case Na and K were then determined simultaneously. Samples for Cl determination were not ashed but directly assayed by the second method of Ramsey et al. [ 151. The chemical contents of the tissues are expressed as the mean in mmol/kg dry wt + S.E.M. (N-number of cochleae assayed). Usually, 3-5 samples were analyzed from each cochlea. Representative cochleae were fixed with 1% osmium tetroxide in dichromate buffer, embedded, stained with toluidine blue, and examined histologically after the EP had been reduced maximally by each treatment. RESULTS Potential and resistance The control medium could be perfused for 120 min with no significant change in the

EP (less than 5 mV). When K-free medium was perfused, the EP typically began its decline within 1.5 min and continued to decline for 30-60 min to a negative value (- 11.9 + 2.5 mV (N = 8); Fig. 1). Superimposed anoxia always led to a negative value typically seen in this condition (about -30 mV). The input resistance of Scala media did not decrease during K-free perfusion (Fig. l), but rather increased slightly. Since the positive component of the EP is dependent upon Na/K-ATPase activity [S], and since an absence of K is expected to inhibit this activity in a way similar to ouabain intoxication, control solution containing 1 mM ouabain was perfused under identical conditions for comparison to the response to the K-free medium (Fig. 1). The decline of the EP was clearly more rapid and to a lower level. Clemistry Although perilymphatic

perfusion of control medium produces no changes of cochlear potentials, a moderate but significant (P < 0.025) drop of the K content of SV was observed inspite of normal Na and ATP content (Figs. 2, 3). The cause of this drop is presently unexplained. The K content of the spiral ligament showed a similar but statistically not significant decline (P > 0.05). The K content of the spiral ligament under normal conditions (non-perfused), and during perfusion with control, K-free, and ouabain-contaming media is shown in Fig. 3. Both perfusion of K-free and ouabain-containing media reduced the K content significantly (P < 0.001) below that of the control perfusion. Fig. 2 shows the effect of the same treatments on the K content of the stria vascularis. Kfree perfusion reduced the strial K content by 34% of control (P < 0.005) while. 1 mh4 ouabain reduced it by 77% (P < 0.001).

152

RESISTANCE (Ktlohn)

11.7

12.7

t

t

60

40

20

0 -20

\ \

\

\

20 \

30

I 40

I

I 60

TIME (Min)

-40

Fig. 1. Response of the second turn EP to perilymphatic perfusion of K-free medium (solid curve) and of 1 mM ouabain (broken curve) after a 10 min perfusion of control medium. Ischemic anoxia was induced after the EP had been maxhnaliy reduced by the K-free perfusion. The average negativity observed (- 12 mV) after K-free perfusion was apprecisbly greater than the example shown here. The resistance between Scala media and the neck during K-free perfusion is shown at the vertical arrows at the top of the figure.

Na content of the stria under the same experimental treatments is indicated in the of Fig. 2. The Na content complements the K content to yield an approximately constant total major-cation content. The Cl content of non-perfused stria was found to be 229 + 25 mmol/kg.dry wt (N = 4). Fig. 3 shows that the ATP content of stria vascularis of cochleae perfused with control and ouabain solutions were in the normal range [ 181, while that of K-free cochleae was reduced 26% (P < 0.001). The

histogram

Histology Following treatment with ouabain the thickness of the stria increased and numerous, large vesicles were noted in the marginal cells (Fig. 4A). There was no noticeable extracellular edema. The amount of swelling of the marginal cells, as determined by the amount of scalloping to the endoiymphatic surface, was somewhat less than that recently observed by Bosher [ 1] . There was little or no swelling at the conclusion of K-free perfusion (Fig. 4B). No scalloping of marginal cells could be discerned and the vesicles were few and small. Again, no extracelhrlar edema was noted. The K-free perfusion additionally caused the strial cells to lose their differential staining properties (the latter effect has

153 800

soo1wl

POTASSIlM

(stria)

(Stria)

t 600

600

t

;

r ii”

2 y” 400

400

y" ;

;

t

t 200

200

0 lronU1 (N-6)

tontml (I.61

K-Free (N-5)

b

OU4blll (N-3)

0 Norm1 (N.11)

Control

K-Free

Fig. 2. Potassium and sodium contents of rtria vascularis after surgery but without perfusion (Normal), after control perfusion (Control), after K-free perfusion (K-free), and after perfusion of 1 mM ouabain (Ouabain). Vertical bars represent standard deviation.

also been noted after arterial perfusion with hype-osmotic artificial blood (Thalmann, R. and Bohne, B.A., unpubl. obs.). DISCUSSION

Central to the discussion are the data on ion content of the SV. Since such meaaurements have not been presented in detail before (see [9) and [ 141 for preliminary accounts) we first examine these results. Ion content of SV The tissue ion content of lyophijized stria *, although not as meaningful a parameter as

intracellular ion activities, is of use in evaluating the effects of experimental treatments. l We did not analyze organ of Corti nor Reissner’s membrane because prelimimuy experiments showed that extrrcellulnr contamination obscured the intracclhdar contribution from these tisauer; they both have large surface to intracell~ volume ratios, while the strip ia relatively ‘dense’, except for its vascular space. In addition, extracellular freeze dried residue containing salts can readily be removed from the endolymphatic surface of the Ma.

154 POTASSIUM (Spiral Ligament)

ATP (Strial)

15

600

t 10

400

r

=

t 6

6

z

y"

i

2

B

F 200

5

0

C Normal (N=6)

Contml (~-6)

K-Free (N=5)

Ouabain (N-3)

Normal (N-4)

contl (N=;

K-Free (N.4)

Ouabrin (N.3)

Fig. 3. Adenosine triphosphate content of stria vascularis and potassium content of under the same four conditions of Fig. 2.

spiral ligament

Other techniques of estimating intracellular ion activities or concentrations either are not readily applicable to strial tissue or have equally many problems in interpretation. (a) Inclusion of extracellular markers, such as inulin, with perfusion fluids is problematic since there are four morphologically identifiable compartments: endolymphatic, perilymphatic (spiral ligament fluid), vascular and intercellular spaces. The distribution of perilymphatically applied inulin throughout the strial extracellular space would likely be far from uniform. (b) Intracellular potential measurement of strial cells is as yet an unestablished technique; diametrically opposite results have recently been reported by two teams of investigators [ 13, 141. The use of ion-selective electrodes would further compound the difficulties involved. (c)‘Electron microprobe and laser mass spectrometric measurements of lyophilized or frozen hydrated tissue potentially have great promise, but also have considerable inherent uncertainties. For instance, tissue preparation may alter the tissue contents before analysis. In order to obtain adequate morphologic preservation for the laser microprobe, the stria and spiral ligament are removed from the cochlea for freezing (Morgenstern, C.E., pers. comm.). It is quite conceivable that significant ion shifts occur during this brief short-circuit period. Any measurements of diffusible Substances, such as ions, which depend on lyophilized tissue are subject to possible shifts of material during freezing and drying. This factor does not constitute a significant problem in our whole tissue measurements. Using a preparation method similar to ours, Ryan and coworkers [17] have shown by X-ray analysis that the elemental boundary between stria and spiral ligament is very sharp. Our primary means of assessing possible endolymph contamination is the measurement of tissue Cl which is at a high level (130 ml@ in endolymph and typically much lower intra-

155

Fig. 4. Photomicrographs of second-turn stria vascularis and spiral ligament after (A) perilymphatic perfusion of 1 mM ouabain. The initial EP was + 80 mV and subsequently declined to - 40 mV during 27 min of perfusion. (B) Comparable region of another cochlea after 35 min of K-free perfusion, during which the EP declined from +84 mV to - 15 mV. E, endolymphatic space; M, marginal cell; I, intermediate cell; B, basal cell; C, capillary; SPL, spiral ligament.

156

cellularly. Our Cl value compares favorably with other tissues [22], thus supporting our contention of minimal ion movements into or out of the stxia during preparation. Although the cation content of stria is not directly comparable to that of other tissues, a qualitative comparison is useful to gain some confidence in the measurements, Data on cation content of kidney and brain from many investigators have been compiled [4,22], The ratios of K : Na and Na : Cl for non-perfused stria are 1.9 and I .2, respectively. The analogous ratios for rat brain are 2.2 and 1.4, while for rabbit kidney tubules they are 1.6 and 1.3; thus the order is consistently K > Na 3 Cl for the three tissues. Even though the water content and the vascular and extracellular spaces were not directly measured in these experiments, we can arrive at an estimate of the average intracelIuiar ion content of eon-perfused stria from our whole tissue data by assu~ng reasonable values for these parameters*: K = 127 mM, Na = 48 mM and Cl = 39 mM. The converted Na value is very close to that of Morgenstern et al. [14] for marginal cell Na (42 mM), and the Cl value compares favorably to the activity determined for other epithelial cells by ion-sensitive electrode (e.g., 35 mM for rabbit gallbladder [Z] ; although it is lower than corresponding estimates of Cl concentration (see [Z] and [22 1). Our K value, although quite normal for many cell types, is substaRti~ly higher than that determined by Morgenstern and coworkers for the superficial luminal layer of the stria (98 mM). There are several possible sources of this discrepancy: (1) There may be large differences in K-content of the different strial cell types. (2) There could be conta~nation from endolymph on our samples, although this is not likely since the samples were thoroughly cleaned of dried endolymph residue before analyses and since our Cl values are not unusually high. (3) 0 ur values depend strongly on the estimated water content of the tissue. (4) Possible problems referred to earlier with the interpretation of the measurements by Morgenstern and associated could also lead to the discrepancy. In particular, if a luminal K pump is present in the stria, even a brief short-circuit period before freezing could reduce the cellular K content. Compa~son of the K content of spiral ligament and stria vascularis is not me~ngf~ since the cellular density of spiral ligament is far lower; much of this tissue consists of extracellular fibrils. The primary value of the spiral ligament data is to ascertain the effectiveness of the perfusions. The pronounced changes observed with ouabain and Kfree perfusion demonstrate that the perfusates reached the basal surface of the stria. The ion content of the stria after ouabain or K-free perfusion cannot readily be converted to average intracellular concentrations because of the observed swelling. Swelling is expected with ouabain intoxication, since most animal cells depend on Na/K-ATPase activity volume regulation. The lesser swelling of the stria as a whole, and the marginal cells in particular, during K-free perfusion suggests a possible difference in the mode of inhibition of EP generation. On the other hand, it may merely represent the difference in extent of inhibition. K-free vs. ouabain The initial changes in the EP during K-free perilymphatic

perfusion closely follow

* Vascularspace = 15%; Hematocrit = 43%; red blood cell ion concentrations = K 107 mM, Na 24 mM, and Cl SO mM; extracellular space = 5%; serum and extracellular ion concentrations = K 5 mM, Na 150 mM, CI 110 mM; tissue water = 80%.

157

those reported by Konishi and Kelsey [6]. These investigators demonstrated the decline of the EP, but chose to observe the reversibility of the effect after only a partial decline (to about +20 mV). During our K-free perfusions the EP did not decline as rapidly nor obtain as low a level as in ouabain (1 mM) intoxication. Since ouabain inhibits Na/KATPase activity by competition with K, one might expect that both treatments would produce closely similar results. Several hypotheses can be advanced to explain the discrepancy : (1) Since all cells require K to maintain general metabolic health, the decline of the EP during K-free perfusion might be due to a massive increase in conductance of the cochlear duct, rather than a specific, inhibitory effect on the EP generator of the stria vascularis. The measurements of input resistance and the fact that the EP turns negative argue against this proposition*. The slight increase in overall resistance observed during K-free perfusion is probably due to removal of the primary current-carrying ion from K-selective pathways bounding the cochlear duct. (2) The two treatments do act in the same way and at the same sites, but ouabain would build up at the site of action to toxic levels faster than K is reduced to inhibitory levels by perfusion. The spiral ligament produces a large unstirred layer so that the movement of ouabain to, or K from, the stria depends predominantly on diffusion rather than on the bulk flow of perfusate. This unstirred layer, coupled with the sensitivity of strial Na/K-ATPase to doses of ouabain much lower than those employed here [8], support this hypothesis. Thus, the time it takes for extracellular K at the stria to drop from 5 mM to a given inhibitory level (e.g., the Zse for strial Na/K-ATPase is 0.9 mM) during K-free perfusion may be appreciably longer than for the ouabain concentration to build to the same inhibitory level (I,, = 3.2 X 10 -3 mM) during perfusion of 1 mM ouabain . The lesser loss of strial K during K-free perfusion could therefore be attributed to the lesser inhibitory action; the lowered level of ATP probably is indicative of another type of damage. This loss of ATP is not, however, sufficient to account for the entire reduction of the EP. When the strial ATP is reduced by the same amount by anoxia as by K-free perilymphatic perfusion (-26%), the EP is still at a positive level of about 50 mV

[181. (3) Ouabain and K-free media act at different sites in the SV. Ouabain may affect all three cell types of the SV, even though the drug concentration may be reduced at the marginal cells due to removal by the blood circulation. This hypothesis is supported by the complete abolition of the +EP by ouabain and the substantial loss (77%) of K. The normal ATP level suggests (as previously reported [ 111) that ouabain primarily interferes with energy use rather than with generation. K-free perfusion, on the other hand, does not lower the EP nor strial K-content to’the same extent as ouabain, but it does lower the strial ATP level significantly. The reduction of ATP suggests interference with cellular energy generation in excess of its interference with energy utilization (EP generation). Since the percent strial K-loss is quantitatively similar to the percent strial ATP-loss, * Although our measurements of resistance do not represent that of the cochlear duct alone, any substantial change in impedance of the cochlear duct should be readily detectable.

158

and since the K-free medium reaches the strial basal cells (as evidenced by the lowered spiral ligament K-content), one might speculate that K-free perfusion primarily damages the basal cells. One could postulate further that the basal cells cooperate with the marginal cells in producing the tEP. Under this theory, the K supplied from the vasculature to the marginal cells is not involved directly with transmural transport. Vascular vs. perilymphatic potassium An adjunct to this third theory postulates compartmentation of K. Vascular K maintams the basal K content of the strial cells while perilymphatic K is utilized for the transepithelial current. The route of the K-current might be basal cells, gap junctions [3], marginal cells*. Additional support for this notion is offered by the measurements of Konishi and coworkers [5] of the accumulation in Scala media of perilymphatically applied radioactive K. Alternatively, one might argue that K-free artificial blood passing through the stria could conceivably have acquired enough K from perilymph while passing above Scala vestibuli to support EP generation from the vascular supply (or in the experiments of Konishi et al., the normal blood may have picked up labeled K from the perilymph). This argument, however, is discounted by the observation [5] that no more labeled K is accumulated in Scala media when applied to Scala vestibuli than to Scala tympani. A parallel in a Na-transporting epithelium, frog skin, is available from electron microprobe analyses which also suggest a syncytial transport process among the cell layers

1161. Preliminary experiments substituting Rb for K in perilymph suggest that the 34% strial K loss observed during K-free perfusion represents the total fraction of strial K in communication with the perilymph. In four ears perfused perilymphatically with Rb-substituted medium the stria contained 39% Rb (i.e., 100 X Rb/(Rb t K)) with the total Rb +K= 547 mmol/kg dry wt. Rb has previously been shown to be an adequate substitute for K in generation of the EP [lo, 201. When considering the results of these K-free perilymphatic perfusions and the previously reported K-free vascular perfusions [20] , one may wonder if the vascular K supply is necessary for maintenance of the EP. The vascular K-free perfusion causes no drop in EP for at least 20 min and then only a slow decline (1 mV/min). This delayed response may be due to slow reduction of perilymphatic K by the vasculature which in turn lowers the EP as seen in the present experiments. However, when dual perfusion was performed with K-free blood substitute and K-containing artificial perilymph, the EP still dropped by about 50 mV in 1.5-2 h (Kambayashi, unpubl. obs.). Even if vascular K is not used directly for EP generation, it seems to be required at least indirectly for the well being of the stria. Further experiments are needed to see if the stria is actually damaged metabolically under these conditions of dual perfusion and to what extent K is lost from the stria. ACKNOWLEDGEMENTS

We are grateful

to Dr. Barbara Bohne for preparing

and interpreting

the histologic

* Intermediate cells are not discussed here since they do not constitute a continuous layer. This does not imply, however, that they do not perform a vital role in strial function.

159

material. This investigation was supported by NSF Grant BNS 77-16842 and NIH Grant NS 06575. REFERENCES [ 1] Bosher, S.K. (1980): The effects of inhibition of the strial Na-K-activated ATPase by perilymphatic ouabah in the guinea pig. Acta Otolatyngol. 90,219-229. [ 21Duffy, M.E., Turnheim, K., Frizzell, R.A. and Schultz, S.G. (1978): Intracellular chloride activities in rabbit gallbladder: Direct evidence for the role of the sodium-gradient in energizing ‘uphill’ chloride transport. J. Membrane Biol. 42, 229-245. [ 31 Jahnke, K. (1975): The fine structure of freeze-fractured intercellular junctions in the guinea pig inner ear. Acta Otolaryngol., Suppl. 336,20. [4] Katzman, R. and Pappius, H.M. (1973): Brain Electrolytes and Fluid Metabolism, p. 114. Williams and Wilkins, Baltimore. [S] Konishi, T., Hamrick, P.E. and Walsh, P.J. (1978): Ion transport in guinea pig cochlea I. Potassium and sodium transport. Acta Otolaryngol. 86,22-34. [6] Konishi, T. and Kelsey, E. (1973): Effect of potassium deficiency on cochlear potentials and cation contents of the endolymph. Acta Otolaryngol. 76,410-418. (71 Konishi, T. and Mendelsohn, M. (1970): Effect of ouabain on cochlear potentials and endolymph composition in guinea pigs. Acta Otolaryngol. 69,192-199. [ 81 Kuijpers, W. (1969): Cation transport and cochlear function. Ph.D. Thesis, University of Nijmegen, The Netherlands. (91 Marcus, D.C., Marcus, N.Y. and Thalmann, R. (1979): Replacement of perilymph with K-free media abolishes positive endocochlear potential (guinea pig). J. Acoust. Sot. Am. 65, 313. [lo] Marcus, D.C. and Thalmann, R. (1979): Substitution of rubidium for potassium in periiymph. Arch. Otorhinolaryngol. 224, 155-156. [ 1 l] Marcus, D.C., Thalmann, R. and Marcus, N.Y. (1978): Respiratory rate and ATP content of stria vascularis of guinea pig in vitro. Laryngoscope 88,1825-1835. [ 121 Matschinsky, F.M. and Thalmann, R. (1967): Quantitative histochemistry of the organ of Corti, stria vascularis and macula sacculi of the guinea pig. I. Sampling procedure and analysis of pyridine nucleotides. Laryngoscope 77, 292-305. [ 131 Melichar, I. and Syka, J. (1980): Distribution of DC potentials in the stria vascularis measured in vitro and in vivo. In: International Physiological Congress, Budapest. (Abstract). [ 141 Morgenstern, C.E., Vosteen, K.H. and Arnold, W. (1980): Formation of inner ear fluids - Perrneability of inner ear membranes. In: Proc Sixth Shambaugh lnt. Workshop on Otomicrosurgery and Third Shea Fluctuant Hearing Loss Symp. Editors: G. Shambaugh and J.J. Shea. Strode Publishers, Huntsville, Ala. (in press). (151 Ramsey, J.A., Brown, R.H.J. and Crogham, P.C. (1955): Electrometric titration of chloride m small volumes. J. Exp. Biol. 32,822-829. [ 161 Rick, R., Dorge, A., Bauer, R., Beck, F., Mason, J., Roloff, C. and Thurau, K. (1980): Quantitative determination of electrolyte concentrations in epithelial tissues by electron microprobe analysis. In: Current Topics in Membranes and Transport, Vol. 13, pp. 107-120. Editor: E.L. Boulpaep. Academic Press, New York. [ 171 Ryan, A.F., Wickham, M.G. and Bone, R.C. (1979): Element content of intracochlear fluids, outer hair cells, and stria vascularis as determined by energy-dispersive Roentgen ray analysis. Otolaryngol., Head Neck Surg. 87,659-665. [ 181 Thalmann, R., Kusakari, J. and Miyoshi, T. (1973): Dysfunctions of energy releasing and consuming processes of the cochlea. Laryngoscope 83, 1690-1712. [ 191 Vurek, G.G. and Bowman, R.L. (1965): Helium-glow photometer for picomole analysis of alkali metals. Science 149,448-450. [20] Wada, J., Kambayashi, J., Marcus, DC. and Thalmann, R. (1979): Vascular perfusion of the cochlea: Effect of potassium-free and rubidium-substituted media. Arch. Otorhinolaryngol., 225, 79-81.

160 [21] Wada, J., Paloheimo, S., Thalmann, I., Bohne, B.A. and Thalmann, R. (1979): Maintenance of cochlear function with artificial oxygen carriers. Laryngoscope 89, 1457-1473. [ 221 Whittembury, G. and Grantham, J.J. (1976): Cellular aspects of renal sodium transport and cell volume regulation. Kidney Int. 9, 103-120.