In this study, we sought to clarify the involvement of CFTR in transepithelial chloride conductance at the mouse ocular surface. An ocular surface open-circuit potential difference approach was introduced, which is based on the generation of PDs across tight epithelia from asymmetries in apical and basal membrane ion conductances. Maurice
34 35 and Klyce
36 measured electrical potentials between the body and tear film in rabbit of −15 to −30 mV (tears negative to blood). This original observation of the ability of the ocular surface to generate a PD was expanded on in the present study to characterize electrogenic Cl
− transport processes in living mice by perfusion of the ocular surface with solutions that created ionic gradients and contained transport activators or inhibitors. In contrast to short-circuit current and single-cell microelectrode approaches, which require tissue and cell isolations in vitro, the ocular surface PD approach is technically simple and permits minimally invasive in vivo measurements under physiological open-circuit conditions. Also, ocular PDs produce robust voltage changes that are readily interpreted in terms of specific ion transport processes. HCO
3 − transport contributed little to ocular surface PDs compared with Cl
−, demonstrating the relative importance of chloride channels in ocular surface anion flux.
Several lines of evidence implicate the functional involvement of CFTR in ocular surface Cl
− secretion, including activation of Cl
− secretion by CFTR agonists and reversal of agonist effects by CFTR inhibitors. The incomplete reversal by GlyH-101 may be related to its external pore-blocking mechanism in which reduced extracellular Cl
− concentration and plasma membrane depolarization are predicted to reduce its potency.
27 While agonists were shown to stimulate CFTR conductance, experiments where Cl
− channel inhibitors were added under low Cl
− but in the absence of agonists indicate that CFTR does not mediate most cAMP-independent Cl
− secretion at the ocular surface. Because DIDS is a relatively nonspecific Cl
− channel inhibitor, the partial DIDS effect in these experiments cannot be interpreted to implicate the activity of specific Cl
− channels.
The absence of CFTR agonist and inhibitor effects in CF mice provides additional evidence for CFTR involvement in ocular surface Cl
− transport. One interesting observation was the near-zero hyperpolarization in CF mice after low Cl
− in the absence of CFTR agonists. This observation seems at variance with the conclusion based on CFTR inhibitor studies that only a small fraction of the low-Cl
−–induced hyperpolarization involves CFTR-mediated transport. However, similar results were found in nasal PD measurements in CF mice,
28 29 where it has been proposed that the low-Cl
− response results from Cl
− transport by a CFTR-regulated Cl
− channel, though not CFTR itself. An alternative possibility is that the non-CFTR Cl
− channel responsible for the low-Cl
− response is downregulated in CF mice.
Previous RT-PCR and immunocytochemical studies have detected CFTR in corneal and conjunctival epithelial cell cultures and tissue mounts from rabbit.
12 13 19 However, functional studies on in vitro preparations relied exclusively on nonspecific channel activators (e.g., genistein) and inhibitors (e.g., glibenclamide), and cAMP-elevating agents (IBMX [3-isobutyl-1-methylxanthine] and forskolin). In our in vivo studies, we treat the ocular surface as a single functional unit as in the intact eye, so that ocular surface PDs represent the aggregate of corneal and conjunctival properties. Another reason that we performed measurements on the minimally perturbed whole ocular surface is that mechanical isolation of the two tissues in vivo disrupted epithelial cell integrity. A machined plastic chamber, beveled to create a seal at the limbus and permit selective perfusion and measurement of cornea or conjunctiva, results in a slow, progressive reduction in PD toward zero. Such contact also imposed a stress that may affect cellular transport processes. Further studies are needed to quantify the relative contributions of cornea versus conjunctiva to the ion channel-dependent PD responses recorded in the study and to determine whether increasing CFTR-dependent fluid transport across each tissue could supplement tear formation in states of aqueous deficiency.
Although CF is not generally associated with overt dry eye symptoms, the results of the current study suggest that CFTR provides an important pathway for agonist-stimulated fluid secretion across the ocular surface. Several factors may account for the mild ocular phenotype in CF. Basal CFTR activity may be very low under normal conditions, so that ocular surface hydration does not depend on CFTR function. Also, upregulation of alternative secretory pathways either in tear-producing glands, or within the cornea or conjunctiva, may compensate for impaired CFTR Cl
− channel function. Indeed, calcium-activated Cl
− channels have been found in some models to compensate for CFTR dysfunction.
37
Our results suggest the possibility of using CFTR activators for treatment of dry eye syndrome. Topical application of IBMX, a phosphodiesterase inhibitor/CFTR activator, has been found to reduce the osmolarity of collected tears in humans and rabbits with KCS.
38 39 In our study, CFTR activation produced sustained, large increases in ocular surface Cl
− secretion, with quantitatively much greater time-integrated hyperpolarization than UTP, an activator of calcium-sensitive Cl
− channels. Currently, topical cyclosporin A, which suppresses downstream inflammatory cascades that ultimately compromise epithelial barrier function, is the only drug approved for treatment of dry eye, with UTP analogues currently under development.
21 22
Although this study was focused primarily on mechanisms of Cl
− secretion, ocular surface PD measurements in response to amiloride indicated the involvement of amiloride-sensitive Na
+ channels, such as ENaC, in ocular surface Na
+ absorption. Inhibitors of Na
+ absorption may provide another therapy for dry eye syndrome. The evidence here for the involvement of amiloride-sensitive Na
+ channels in mice contrasts with short-circuit experiments in rabbit tissue.
17 40 However, functional evidence for ENaC was reported in bovine cornea by Midelfart,
41 42 and ENaC protein has been found by immunocytochemistry in rat and human cornea.
43 The roles of individual ocular surface transporters are probably species specific; for example, cAMP activates basolateral K
+ channels in frog corneal epithelium but inhibits them in rabbit.
44 45 The amiloride-induced depolarization was greater in CF than wild-type mice, as has been reported in nasal PD measurements.
28 The greater depolarization in nasal PDs in CF mice has been interpreted as increased ENaC activity in CF resulting from direct or indirect CFTR-ENaC interaction, though it remains controversial whether altered electrochemical driving forces for Na
+ absorption in CF can account fully for the greater depolarization in CF.
46 47
In summary, measurements of ocular surface PD in mice indicated the involvement of CFTR in ocular surface Cl
− secretion. Based on evidence for Na
+ absorption by corneal and conjunctival epithelia
1 2 17 and the amiloride effects described herein, ocular surface PDs should also be useful in studying mechanisms of electrogenic Na
+ transport at the ocular surface. The substantial and sustained activation of CFTR-mediated Cl
− secretion by CFTR agonists supports the evaluation of CFTR-selective activators as potential treatment for dry eye syndromes.
The authors thank Liman Qian for mouse breeding and genotype analysis and Luis Galietta, Yaunlin Song, and Danieli Salinas for advice on electrophysiology.