The activation of chloride transport by epinephrine ...

The activation of chloride transport by epinephrine and Db cyclic-AMP in the cornea of the rabbit S. D. Klyce,* A. H. Neufeld, and J. A. Zadunaisky

The addition of epinephrine to the solution bathing the epithelial side of the isolated rabbit cornea produced a biphasic increase in the short-circuit current with little or no change in the potential difference. Addition of aminophylline or theophylline, produced prolonged stimulation of the current, presumably through inhibition of phosphodiesterase. Db cyclic-AMP produced a rapid and sustained increase of the current as well as a large increase in the net transport of chloride ions from endothelial to epithelial side. This chloride pump, activated by an elevation of cyclic AMP, is located in the epithelium of the rabbit cornea. In Cl-free solutions no activation by these compounds of the remaining sodium current was found. The cyclic-AMP content of corneas incubated in the presence of theophylline and epinephrine was significantly increased. It is concluded that epinephrine influences the function of the rabbit corneal epithelium by increasing the level of cyclic AMP which activates a chloride pump operating toward the tear side. The implications of the presence of this chloride pump are discussed.

G

exert their action through the cyclic-AMP system, epinephrine is an agent with well known and potent action on the eye.2 Recently, Chalfie, Neufeld, and Zadunaisky3 demonstrated that epinephrine has a stimulatory effect on the transport function of the frog corneal epithelium. This present work reports the activating effect of epinephrine, Db cyclic-AMP, and inhibitors of phosphodiesterase on the transport functions of the epithelium of the rabbit cornea. The stimulation of the short-circuit current (S.C.C.) of the rabbit corneal epithelium by epinephrine and cyclic-AMP derivatives consists of an activation of a chloride transport located in the epithelium and directed toward the tear side. This chloride pump has not been reported before in the mammalian cornea. Its presence was unmasked during a study of the action of epinephrine

'yclic-AMP has been shown to be the "second messenger" for a number of hormones and pharmacologic agents and to influence a variety of cellular functions.1 Among the better studied hormones that From the Department of Ophthalmology and Visual Sciences and Department of Physiology, Yale University School of Medicine, New Haven, Conn. 06510. Supported by United States Public Health Service, Research Grants EY-00382 and EY-00758. Manuscript submitted Aug. 17, 1972; manuscript accepted Nov. 6, 1972. Reprint requests: Dr. J. A. Zadunaisky, Department of Ophthalmology and Visual Science, Yale University Medical School, 333 Cedar St., New Haven, Conn. 06510. * Present address: Division of Ophthalmology, Stanford University Medical Center, Stanford, Calif. 94305.

127

128 Klyce, Neufeld, and Zadunaisky

Fig. 1. Chamber utilized to mount rabbit corneas as a membrane. P.D.: Potential difference agarRinger bridge A: Air inlet. V: Vacuum connector to suction groove. S: Stirring bar. W: Water jacket for temperature control. on the short-circuit current of isolated rabbit corneas. Methods Eyes were enucleated from 10 to 12 pound New Zealand White rabbits after asphyxiation with CO2 or killing by a blow on the head. The cornea was dissected out and mounted in a chamber depicted in Fig. 1. The dissection of the cornea required some 5 to 10 minutes. The epithelium was frequently washed with Ringer fluid to prevent damage through evaporation. The conjunctiva was carefully dissected away from the limbal area so that it would not plug the vacuum mounting ring of the lucite chamber. The eye was then lifted with forceps by the optic nerve and carefully lowered onto the tear side hemichamber where the vacuum ring held it firmly in place, without applying pressure or strain to the cornea. The posterior segment of the eye was then carefully sliced off with a razor blade exposing the .lens. This was removed by tilting out the cortex, leaving the capsule attached to the zonules. The zonules and the iris were removed by gentle detachment so that the vacuum was not broken. The aqueous humor was replaced with Ringer fluid by alternate rinsing and aspiration. The aqueous side of the chamber was then pressed lightly in place and the chamber was mounted on a stand. Subsequently, 2.6 and 2.0 ml. of Ringer fluid were added to the aqueous and tear sides, respectively, the electrical connections made, and the stirring begun. Commonly, this careful mounting technique led to resting corneal resistances of 8 K^cm.2 or more, reducing passive ion fluxes to low levels. In most cases corneas failing to generate 15 mV in the first 30 minutes of incubation were discarded. The double-sided lucite chamber, depicted in Fig. 1, is similar to chambers used previously for other membrane transport studies,4"7 but adapted to rabbit corneal dimensions and mounting requirements. The clamping ends of the hemichambers were machined with a radius of curvature of 8.5

Investigative Ophthalmology February 1973

mm. The concave end contained the vacuum mounting ring forcing the perilimbal sclera to adhere, thus stabilizing the tissue during the final dissection. Water jackets surrounded each hemichamber to keep corneas at a normal temperature of 34° C. Stirring was accomplished with small magnetic fleas at 200 r.p.m. and by the constant aeration provided by air bubblers. Conventional agar-Ringer bridges were used for measuring corneal potential and for applying S.C.C. by means of an automatic voltage clamp. The 18 ficm.2 of chamber resistance was generally uncompensated in these experiments as corneal resistance rarely fell below 2 K^cm.2. The spherical portion of the cornea exposed, to the fluid in this chamber was 1.00 cm.2. The Ringer fluid used in these experiments contained the amounts of Na+, K+, and Cl~ normally found in rabbit aqueous humor and was similar to that used previously.7 For the long-term experiments described here the buffer capacity of the Ringer was increased through the incorporation of TRIS-PO... This was prepared by titrating 100 mM TRIS with l.ON H3PO4 to pH 7.4. When required, Cl-free solution was prepared by SCV2 substitution. The final composition of the normal Ringer was 103.4 mM NaCl; 20.3 mM Na:SO.,; 2.2 mM K2HPO4; 0.5 mM KHoPCX; 5.24 mM H3PO4; 0.61 mM MgSCX; 0.7 mM calcium gluconate; and 20 mM Tris (hydroxymethyl) aminomethane. The final osmolarity of solutions was adjusted to 305 mOsm. by the addition of sucrose. To replace solutions entirely in the chambers as with the sulfate experiments, the tear-side solution was aspirated prior to exchanging the aqueous side to prevent corneal buckling from the hydrostatic pressure. Each side was exchanged at least twice with the new solution. Drugs were added 60 to 90 minutes after the S.C.C. had stabilized. This was done by withdrawing a sample of chamber fluid to dissolve the drug, then replacing it. The following compounds (Sigma Chemical Company, St. Louis, Mo.) were used: N6, O2'-dibutyryl adenosine 3':5'-cyclic monophosphoric acid (Db cyclic-AMP) as the sodium salt; adenosine 3': 5'- cyclic monophosphoric acid (cyclic-AMP); 1-epinephrine bitartrate; aminophylline; theophylline, and hypoxanthine. Unidirectional radioisotopic fluxes were measured under short-circuit conditions. To prepare the 36C1 Ringer, NaOH was used to neutralize the H3GC1 to pH 7, then the resultant Na3CCl was substituted for an equivalent amount of Ringer NaCl. The specific activity of the radioactive Ringer was 29 /tCi per milliequivalent of CH Ringer containing 36C1 was added to the appropriate side when the corneas were mounted. Sampling of the hot side was carried out three times during the course of the experiment with a

Activation of chloride in rabbit cornea 129

Volume 12 Number 2

s.c.c. E 6

u o 4 o

40

/-EPINEPHRINE BITARTRATE \0'7 M P.D. ooooo 0 oo 0 0 o o Oo

n

OOoOOOOOOOOoo0000 0o0o000 0°OOOgoOoOooOooO

» oooooooooooo 0 o o

30 C 20 10

3

4

5

6

TIME (hr) Fig. 2. Effect of epinephrine on the S.C.C. and potential difference of the isolated rabbit cornea. 10 /A constriction micropipette. Every 15 minutes a 250 /xl sample was taken from the less radioactive side and replaced with the same amount of fresh Ringer. The corneal potential was measured for two seconds after each sample was taken. Unidirectional fluxes were calculated according to methods previously employed.0 The Cl~ fluxes reached steady state within 45 minutes of incubation. Data reported here includes only fluxes obtained after this stabilization period. Cyclic-AMP levels were determined on a total corneal extract by a method similar to that described by Brown and co-workers.8 Determinations were made on corneas obtained from eyes frozen in liquid nitrogen and pre-incubated corneal buttons of a diameter of 10 mm. The corneal buttons were incubated for 60 minutes in normal Ringer at 34° C. after which theophylline was added at a final concentration of 10-3M. Three minutes after the addition of theophylline, epinephrine at a final concentration of 10-°M was added for the final minute of incubation. All corneas were rapidly placed in boiling KOH to denature the protein prior to running the assay. The cyclic-AMP level in each cornea was determined in triplicate.

Results Electrical characteristics of the rabbit corneal preparation. After one hour of incubation, resting corneal potentials of up to 40 mV were observed th© S.C.C stabilized at values from 2 to 5 ^A per square centimeter. Corneal resistances were highly dependent upon the success of the mounting procedure, but were greater than 5

KQcm.2, in some cases exceeding 14 KQcm.2. It is possible to infer from these results that corneal resistance in the living rabbit eye may be 15 KQcmr or more with an intact epithelium. Sustained S.C.C. had a depolarizing effect on the corneal potential. If the circuit is opened as shown in the last period of Fig. 4, the corneal potential generally increased spontaneously by about 35 per cent. Such an effect is most probably similar to the long, current-induced transients observed in frog gastric mucosa potentials9 and can be explained in terms of an intracellular redistribution of ions. Epinephrine response of the cornea. The addition of 1-epinephrine bitartrate to the tear side solution induces a rapid and biphasic response in corneal S.C.C. (Fig. 2). The onset of the response occurred after some 15 seconds. Total mixing time of the chamber was 3 to 4 seconds, hence, the short diffusional delay observed strongly suggests that the target is the epithelium. Increases in S.C.C. were observed at concentrations as low as 10~9M with an apparent ED50 of 5xlO~8M. Values for the effect of epinephrine on the electrical measurements of the rabbit cornea are presented in Table I. Corneal response to theophylline and


Table I. Effect of 10~7M 1-epinephrine bitartrate on rabbit corneal electrical properties l-Epinephrine bitartrate (10~7M) Initial

Peak

Per cent change

3.54 t 0.42 7.51 ± 0.39 S.C.C. P.D. 22.9 ± 3.40 26.8 ± 3.20 R 7.00 ± 1.21 3.70 ± 0.57 Means (± S.E.) for six comeas. The dose was the tear side only. The changes in current and were significant (p < 0.01).

112 -47 added to resistance

aminophylline. At a final concentration of 10~3M in both sides of the chamber, theophylline induces a biphasic response in the S.C.C. This is composed of a sharp increase followed by a sustained stimulation of the current which reaches twice control levels (Fig. 3). In five of the corneas treated with theophylline there was also a slight increase in corneal potential. Aminophylline at the same concentration produced a better response with a time course similar to that of theophylline. Average results for the methylated xanthines are presented in Table II. Fig. 4 shows the response to theophylline (10~3M) and the subsequent response to epinephrine (10~7M). While the responses sum, one on the other, they most likely lead to the saturation of a single process. When the biologically inactive hypoxanthine was used, there was no response in the corneal S.C.C. (Fig. 5). However, the same cornea responded to the subsequent addition of theophylline. Corneal response to Db cyclic-AMP'. The addition of Db cyclic-AMP to both sides bathing the cornea led to a gradual but sustained increase in the S.C.C. (Fig. 6). The response to this compound was statistically greater than that of maximum doses of both phosphodiesterase inhibitors and epinephrine. The 5 to 10 minute delay in the onset of response was probably due to poor penetration. Because of the large and sustained type of response, Db cyclic-AMP was chosen for the determination of ionic flux.


Action of Db cyclic-AMP on chloride fluxes. Fig. 7 shows the electric response to Db cyclic-AMP as well as its effects on the aqueous to tears Cl30 flux. The main effect consists of an increase in the S.C.C. with a matching increase in the C13G flux from aqueous to tear side. A slight increase in corneal potential, about 15 minutes after the addition of Db cyclic-AMP, was of common occurrence; yet, during the stable portion of the response the potential was significantly decreased. A large decrease in corneal resistance accompanies the increase of the aqueous to tears flux of chloride, while the increase in the passive C13G flux in some of the experiments was negligible. Data for the values of the C13G fluxes are presented in Table III. All of the net fluxes were significantly different from zero. The unidirectional chloride fluxes were fairly stable for up to four hours in untreated corneas, with a slight tendency to increase with time. The result of stimulation with Db cyclic-AMP is the occurrence of a net chloride transport outwards. Sodium fluxes. Table IV lists the results of experiments in which the unidirectional sodium fluxes were measured. The data includes flux periods measured during the second hour of incubation comparable with the control period in the C13G fluxes. The level of S.C.C. for these corneas was statistically similar to the level found in the CPG fluxes. The unidirectional tears to aqueous sodium flux, which is in the direction of its active transport was 80 per cent of the average S.C.C. (0.106 /xEq/cm.2hr.). The corneal sodium permeability calculated from the passive fluxes ranged from 2.5 to 4,3 x 10~4cm. per hour. In an experiment shown in Fig. 8, the addition of 10~3M Db cyclic-AMP did not significantly alter the tears to aqueous unidirectional sodium flux. C\~-free experiments, In Cl^-free Ringer the response of the S.C.C. to Db cyclicAMP was markedly reduced as shown in Fig. 9. A normal response was obtained, however, when Cl~ was reincorporated to the bathing solution. Further addition of


Volume 12 Number 2

THEOPHYLLINE I0" 3 M

,s.c.c.

E 6 u

- 20 Q

2

- 10 n

1

2

3 TIME (hr)

4

Fig. 6. Sustained response doubling the values at control of the S.C.C. of the rabbit cornea, produced by Db cyclic-AMP.

the S.C.C. are associated with increases of cyclic-AMP levels in the tissue. This is supported by the rapid and sustained increase in the current obtained with the more penetrating Db cyclic-AMP derivative and by the actual increase of cyclic-AMP determined analytically on incubated corneas in the presence of these compounds. When chloride ions are replaced by sulfate, Db cyclic-AMP evokes no response and the response to epinephrine is small. Since under these experimental conditions the current probably results from the transport of sodium ions from the epithelial to the endothelial side, it appears that epinephrine and Db cyclic-AMP have little effect on the sodium current. In fact, the unidirectional flux of sodium from epithelium to endothelium did not change significantly during the increase in total electrical current produced by Db cyclicAMP. Further study of this observation seems indicated. The net flux of chloride, on the other hand, was very much increased by Db cyclic-AMP during the increase in S.C.C. The change in the total electric current and the increase in the net chloride flux were statistically similar. Therefore, it may be assumed that most of the corneal current was due to activation of the chloride

transport from the endothelial to the epithelial side. This resulted in an increased unidirectional flux toward the tear side with only a small increase in the passive flux of chloride from the epithelial to the aqueous side. It is important to mention that all the data for fluxes presented here correspond to measurements obtained after equilibration of the radioactive species with the stable ions in the cornea. It is well known that with flux data obtained before equilibrium is attained the observations are not valid and lead to unusual conclusions.21 Chloride also appears to move toward the tears under resting conditions, that is, before activation with cyclic-AMP-related compounds. However, net chloride outwards flux at rest is very small, even though statistically significant. Net sodium flux under resting conditions was in the direction epithelium to endothelium, as previously demonstrated by Dorm, Maurice, and Mills.5 As in the case of these authors, however, not all of the current moving across the isolated rabbit cornea can be accounted for by net sodium transport. A fraction of the current is apparently then carried by chloride ions moving outward. This fraction is small and not readily detectable. Perhaps the reason why we were



30

60

90

120 150 TIME (min)

180

210

240

Fig. 7. Increase of the unidirectional flux aqueous to tears (bars) and simultaneous increase in S.C.C. of rabbit cornea produced by Db cyclic-AMP. Note that the lack of effect in the potential difference is accompanied by a great reduction of the resistance. The increase in net chloride flux and in current was statistically similar.

able to detect it in our experiments is that we were able to obtain higher electrical resistance, and therefore, better ion characteristics than was true in earlier studies. Nevertheless, the convincing reason for accepting the existence of chloride transport is the sizable increase in net chloride movement induced by an increase in cyclicAMP. The mechanism activated by Db cyclicAMP and epinephrine appears to be located in the epithelium. Corneas deprived of endothelium showed the expected electrical

responses of epithelium and responded to drugs as well as intact corneas. Corneas deprived of epithelium, on the other hand, had very low resistances and did not respond to any of the tested drugs. The effect of cyclic-AMP on the epithelial cells appears to be related to a change in permeability with a secondary increase in pump activity. The rapid rise in current is not accompanied by changes in the potential difference in most of the experiments presented here in the rabbit cornea as in the frog cornea.3 The consequent de-

Volume 12 Number 2


Table III. Effect of 10~3M Db cyclic-AMP on rabbit corneal electrical properties and chloride transport S.C.C. (fiA/cm.2) Control period Treated period Per cent change

P.D. (mV)

R (Mem.2)

Chloride fraction (fiEq/cm. • hr.) (nEq/cm. • hr.) f/iEq/cm* • hr.) of S.C.C. rci

JTA

TCI 1 AT

2

2

2.80 ±0.01 24.0 ± 0.20 8.61 ±0.11 0.044 + 0.002

0.086 ± 0.004

0.042 ± 0.005

0.40

6.84 ± 0.07 22.2 ± 0.30 3.32 ± 0.08 0.066 ± 0.004

0.260 ± 0.004

0.194 ±0.006

0.76

144 -61 50 -8 362 202 Means (± S.E.) of six aqueous to tears and six tears to aqueous fluxes. The hour prior to treatment (control period) is co pared to an hour during the stable response (treated period). The difference between these periods are significant (p< 0.0

Table IV. Control level of sodium transport in the rabbit cornea S.C.C. (tiA/cm.2) 2.85 ±0.10

P.D. (mV) 19.4 ± 0.8

R kSlcm* 6.94 ± 0.30

J TA

(nEq/cm.2 • hr.) 0.084 ± 0.009

JNa

TA'o

(fiEq/cm.2 • hr.) 0.045 ± 0.005

inet

(fiEq/cm.* • hr.) 0.039 ± 0.010

Means (± S.E.) of four samples from each of three aqueous to tears and three tears to aqueous fluxes. The difference between unidirectional fluxes is significant (p < 0.02).

crease in resistance could be due to an increase in the permeability to chloride of one of the cell membranes of the epithelium. The cyclic-AMP content of incubated corneas was lower than that of freshly dissected ones. However, treatment with epinephrine and a phosphodiesterase inhibitor raised the cyclic-AMP content of incubated corneas. Together, these results suggest that if the chloride transport does indeed depend upon tissue levels of cyclic-AMP, then one would expect a higher transport level of chloride in the intact cornea than in the isolated organ. Moreover, since most cyclic-AMP is intracellular, and most of the measured protein is extracellular it would appear that the true cyclic-AMP content of corneal cells may be among the highest in the body. It is reasonable to speculate that in the intact eye, catecholamines derived from the ocular fluids or adrenergic fibers may stimulate adenyl cyclase to sustain these high levels of corneal cyclic-AMP. The manner in which the chloride pump in the rabbit cornea was activated was similar to what Chalfie, Neufeld, and Zadunaisky3 reported for chloride transport of the frog cornea. The frog cornea trans-

Table V. Cyclic-AMP content of incubated rabbit corneas Treated cornea 1 hour +1O-SM

theophylline (3 Control min. ) +10~6M cornea 1 hour epinephrine (1 (pmoles/mg. min.) (pmoles/mg. Per cent protein) protein) increase Animal A B C D E F

5.5

6.5 6.2 9.0

1.9 3.1 3.2

4.8 5.9 6.8

103 48 64 153 90 113

6.5 ±0.5

95 ±14

3.2 4.2

Means 3.6 ± 0.5 (± S.E.)

The increase in cyclic-AMP in the treated corneas was statistically significant (p < 0.01).

ports chloride ions toward the tear side by a pump mechanism located in the epithelium.6' lx> 12 This transport is rapidly activated by epinephrine, theophylline, and Db cyclic-AMP.3 Furthermore, a simultaneous increase in light transmission measured through partially opaque, swollen frog corneas was found when the chloride pump was activated by aminophylline.3 The presence of chloride transport in mammalian corneas described here makes species



- 30

60

90 120 TIME (min)

Fig. 8. Lack of effect of Db cyclic-AMP on the unidirectional flux of sodium from epithelium to endothelium.

10 . S O ; RINGER / \

CYCLIC ObAMP

cr

\

I0"3 M

RINGER

"I

1

1

6 -

/CYCLIC ObAMP IO-SM

/

1

^s.c.c. 4 0

2 0

1

\l o

I

40 30 5 O

° o

°

\o.

0

-

20 o a. 10

1

TIME (hr)

Fig. 9. In Cl--free solutions the action of Db cyclic-AMP is very small. Replacement to Cl containing solution induces great increase in current and no further stimulation with a second dose of Db cyclic-AMP.

differences less pronounced and opens the question of the regulation of corneal hydration to further exploration, as discussed below.

Fig. 11 presents a diagram of the orientation and site of the forces that contribute to keep the mammalian and amphibian cornea dehydrated. In the mammalian cor-


Volume 12 Number 2

S O ; RINGER g |.

SO4 RINGER

Cl* RINGER

CPWEPHRME I0*7 M

40 30 ^

20 2

o.

10

TIME ( M Fig. 10. Reduced response to epinephrine in Cl--free solutions. The spikes correspond to periods of washout. In Cl containing Ringer, the response is much greater than in Cl--free solutions.

ENDOTHELIUM

EPITHELIUM

RABBIT

ENDOTHELIUM

EPITHELIUM

FROG

Fig. 11. Diagram indicating the location and type of pumps of the cornea of the rabbit and the frog. The arrows indicate the direction of transport of the ions.

nea (rabbit), it has been demonstrated that the endothelium is the site of a fluid movement from stroma to aqueous humor. This fluid pump is dependent on metabolic energy and is inhibited by ouabain.13"15 A net transport of sodium could be involved as the primary force in this movement of water described in the endothelium. The absence of a sizeable potential difference across the endothelium speaks against an electrogenic sodium pump and a neutral

mechanism can be postulated as an alternative. The epithelium of the rabbit cornea transports sodium ions inwards.5 This sodium transport is oriented in such a way as to offer very little help in an understanding of how the hydration is maintained under physiologic conditions since ions, and therefore water, will move into the stroma from the tears driven by the sodium pump. In contrast, the presently found transport


of chloride located in the epithelium and directed toward the tear side could well be a mechanism for dehydration of the corneal stroma. Both an ionic pump in the endothelium moving fluid toward the aqueous and/or an ionic pump in the epithelium operating toward the tears, can dehydrate the stroma efficiently. In the frog cornea, the chloride pump of the epithelium is activated by cyclic-AMP and it can control corneal hydration.3 The small sodium pump described10 in the frog corneal epithelium is not activated by cyclic-AMP and it is difficult to demonstrate under our experimental conditions. The endothelium of the frog cornea has not been thoroughly explored but it is electrically silent compared to the epithelium. The action of cyclic-AMP and the adenyl cyclase system on the transport and permeability properties of other biologic tissues has been known since the action of theophylline and cyclic-AMP on the toad bladder was reported.17 Many other systems including frog skinls gastric mucosa,19 and intestine20 have been shown to respond to cyclic-AMP. The stimulation of the chloride transport by cyclic-AMP in the intestine of the rabbit20 is of interest and might indicate a similarity of action of these compounds in the two different epithelia. In summary, the activating effect by epinephrine of the S.C.C. of the rabbit cornea is mediated by an increase in cyclic-AMP in the tissue. The mechanism of activation consists of stimulation of a chloride pump in the epithelium and substantial net chloride transport moving chloride ions toward the tear side. The technical assistance of Tatiana Selinger and Ellen Page are gratefully acknowledged. REFERENCES 1. Robison, A. G., Butcher, R. W., and Sutherland, E. W.: Cyclic AMP, New York, 1971, Academic Press.


2. Davson, H.: The Eye, Volume I, New York, 1969, Academic Press, p. 245-251. 3. Chalfie, M., Neufeld, A. H., and Zadunaisky, J. A.: Action of epinephrine and other cyclicAMP-mediated agents on the chloride transport of the frog cornea, INVEST. OPHTHALMOL.

80: 644, 1972. 4. Ussing, H. H., and Zerahn: Active transport of sodium as the source of electric current in the short circuited isolated frog skin, Acta Physiol. Scand. 23: 110, 1951. 5. Donn, A., Maurice, D. M., and Mills, N. L.: Studies on the living cornea in vitro. I. Method and physiologic measurements, Arch. Ophthalmol. 62: 741, 1959. 6. Zadunaisky, J. A.: Active transport of chloride in frog cornea, Am. J. Physiol. 211: 506, 1966. 7. Klyce, S. D.: Electrical profiles in the corneal epithelium, J. Physiol. 226: 407, 1972. 8. Brown, B. L., Albano, J. D. M., Ekins, R. P., et al.: A simple and sensitive saturation assay method for the measurement of adenosine 3':5'-cyclic monophosphate, Biochem. J. 121: 561, 1971. 9. Kidder, G. W., Ill, and Rehn, W. S.: A model for the long time-constant transient voltage response to current in epithelial tissues, Biophys. J. 10: 215, 1970. 10. Klyce, S. D.: Electrophysiology of corneal epithelium, Ph. D. Thesis, Yale University, 1971. 11. Zadunaisky, J. A., Lande, M. A., and Haffner, J.: Further studies of chloride transport in frog corneas, Am. J. Physiol. 211: 1832, 1971. 12. Zadunaisky, J. A., and Lande M. A.: Active chloride transport and the control of corneal transparency, Am. J. Physiol. 211: 1837, 1971. 13. Mishima, S., Kaye, G., Takahashi, G. H., et al.: The function of the corneal endothelium in the regulation of corneal hydration. In: The Cornea. Macromolecular organization of a connective tissue. Langham, M., editor. Baltimore, 1969, The Johns Hopkins Press, p. 207. 14. Dickstein, S., and Maurice, D. M.: The metabolic basis to the fluid pump in the cornea, J. Physiol. 221: 29, 1972. 15. Maurice, D. M.: The location of the fluid pump in the cornea, J. Physiol. 221: 43, 1972. 16. Candia, O. A., and Askew, W. A.: Active sodium transport in the isolated frog cornea, INVEST. OPHTHALMOL. 7: 404, 1968.

17. Orloff, J., and Handler, J. S.: The similarity of effects of vasopressin, adenosine 3,5 phosphate (cyclic-AMP) and theophylline on the toad bladder, J. Clin. Invest. 41: 702, 1962. 18. Baba, W. I., Smith, A. J., and Townsend, M. M.: The effects of vasopressin, theophylline,

Volume 12 Number 2

and cyclic 3',5' adenosine monophosphate (cyclic-AMP) on sodium transport across the frog skin, Q. J. Exp. Physiol. 52: 416, 1967. 19. Harris, J. B., and Alonso, D.: Stimulation of gastric mucosa by adenosine 3'5' monophosphate, Fed. Proc. 24: 368, 1965.


20. Field, M.: Ion transport in rabbit ileal mucosa II: Effects of cyclic 3'5' AMP, Am. J. Physiol. 221: 992, 1971. 21. Green, K.: Ion transport in isolated cornea of the rabbit, Am. J. Physiol. 209: 6, Dec. 1965.