Receptor-Mediated Activation of a Cl⁻ Current by LPA and S1P in Cultured Corneal Keratocytes

Abstract

purpose. This study was designed to examine the effects of lysophosphatidic acid (LPA) and sphingosine-1-phosphate (S1P) on Cl⁻ currents (ICl_LPA) in cultured corneal keratocytes isolated from the corneas of New Zealand White rabbits.

methods. ICl_LPA and resting voltages were recorded with the amphotericin perforated-patch technique. Phenotype was determined with antibodies to α-smooth muscle actin.

results. Keratocytes cultured in serum have a phenotype (myofibroblast) and ionic currents similar to those of keratocytes isolated directly from corneas during wound healing. LPA and S1P both activated ICl_LPA in a dose-dependent manner, and the LPA receptor–specific antagonist dioctyl-glycerol pyrophosphate (DGPP) blocked the LPA response, but not the S1P response. In addition, a relatively inactive form of LPA (LPA 8:0) was relatively ineffective in activating ICl_LPA. Activation of ICl_LPA significantly depolarized the cells, and this depolarization was reversed by blocking ICl_LPA with 4,4′-diisothiocyanostilbene-2,2′-disulfonic acid (DIDS) or 5-nitro-2-(3-phenylpropylamino) benzoic acid (NPPB).

conclusions. These results demonstrate that activation of ICl_LPA by LPA in cultured corneal keratocytes is receptor mediated and that ICl_LPA can also be activated by S1P. From a functional standpoint, this work confirms that the current, which is typically thought of as purely volume-activated, can be activated through a receptor. In addition, activation of ICl_LPA results in depolarization of the keratocyte. Finally, this work demonstrates that cultured corneal keratocytes can act as a model for the study of ion channel function in keratocytes during corneal wound healing.

After injury, whether it be the result of trauma, surgery (including refractive procedures), or infection, the cornea uses a variety of cellular mechanisms to repair itself. These events include reepithelialization, collagen repair, and repopulation of the stroma and endothelium in damaged areas. This process can be quite effective and often leads to complete healing of the cornea with no associated visual impairment. Unfortunately, the wound-repair process can also result in stromal edema, scarring, ulceration, or, in the case of refractive surgical procedures and penetrating keratoplasty, unpredictable and fluctuating refractive errors. To better understand and possibly to control the wound-healing process, an understanding of the basic cellular mechanisms of corneal wound healing is essential. 

In the corneal stroma, keratocytes are responsible for the bulk of the corneal wound-healing response. Corneal keratocytes reside between the collagen lamellae of the stroma, normally in a quiescent state. Approximately 10% of the dry weight of the cornea is composed of keratocytes, of which there are an estimated 2.43 million per human cornea. ^{1 2} After injury, keratocytes activate and migrate to the wound site, where they proliferate and initiate a host of wound-healing activities related to matrix degradation and repair. During wound healing, corneal keratocytes lose the prominent whole-cell ionic currents typically found in quiescent cells: a K⁺-selective outwardly rectifying current and a voltage-sensitive, tetrodotoxin (TTX)-inhibitable Na⁺ current. ³ These activated, migrating cells also become responsive to serum and lysophosphatidic acid (LPA) by activating a Cl⁻ current (ICl_LPA) that is also activated during cellular swelling. ⁴ This is significant, in that ICl_LPA is rarely seen in keratocytes isolated from uninjured corneas. 

ICl_LPA activation is significant, in that LPA, along with several other LPA-like phospholipid growth factors (PLGFs) including cyclic LPA, alkenyl-LPA, lysophosphatidylserine, and phosphatidic acid, are present in the aqueous humor and/or the lacrimal fluid. ⁵ In a previous study, we found that corneal wounding resulted in a physiologically significant increase in the concentration of these lipids. In this same study, it was shown that LPA, alkenyl-LPA, and lipopolysaccharide LPS all significantly increased cell mitosis in acutely isolated rabbit corneal keratocytes. In other cell types, the biological activities of LPA and PLGFs include mitogenic or antimitogenic effects, regulation of Ca²⁺ homeostasis, regulation of the actin cytoskeleton, and inhibition of apoptosis. ^{6 7}  

To date, there have been nine receptors cloned that are specifically activated by the recently recognized class of PLGFs. These receptors belong to two distinct gene families, including one in the PSP24 family and eight in the LPA-sphingosine-1-phosphate (S1P) family. ^{6 8 9} PSP24 and LPA_1-3 all bind LPA, whereas S1P_1-5 most avidly bind S1P. All the PLGF receptors studied to date have been linked to a G-protein signaling cascade. LPA₁ has been linked to Gα_i, ^{10 11} LPA₂ appears to be linked to both Gα_i and Gα_q, ^{11 12} and LPA₃ appears to be linked only to Gα_q. ¹³ We have published functional evidence that the LPA-activated response in acutely isolated wound-activated keratocytes (WAKs) is receptor mediated. ^{4 7} The evidence from these studies indicates that the receptors mediating ICl_LPA responses in WAKs are activated by both alkenyl-GP and LPA, but not by lysophosphatidyl choline. All these previous rabbit keratocyte electrophysiology studies were performed in acutely isolated cells, whereas the functional studies were performed in cultured cells. The present study was designed in part to determine whether, indeed, cultured rabbit corneal keratocytes also contain ICl_LPA. Because many cell types change their ion channel profiles once placed in culture, this is an important question. If cultured keratocytes contain ICl_LPA, these cells can serve as a model to examine the transduction pathways underlying the receptor-mediated response, as well as the functional significance of ICl_LPA activation in WAKs. This is significant, in that with the exception of its activation in keratocytes, this current is typically found to be activated by increases in cell volume and not through a receptor-mediated pathway. ^{4 7} This study was also designed to determine whether cultured keratocytes have minimal K⁺ and Na⁺ current expression, as seen in WAKs. Finally, the study was designed to determine the effect of activation of ICl_LPA on the resting voltage (E _m) of these cells and to examine the possibility that S1P also activates ICl_LPA. 

Methods

Cell Culture and Immunohistochemistry

Corneal keratocytes were isolated from the corneas of 2- to 3-kg New Zealand White rabbits, as previously described. ¹⁴ Cells were grown using DMEM (Cellgro; Mediatech, Inc., Herndon, VA) and Ham’s F12 (Gibco, Rockville, MD) media plus 14% fetal calf serum (FCS). Cells were passaged at approximately 70% confluence. Passage 2 to 4 cells were used in all experiments, typically within three days of passage. All experiments in animals were performed in conformity with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

To determine the phenotype of the cells we were using for the patch-clamp studies, the method described by Masur et al. ¹⁵ was used. Briefly, cells grown in our standard culture conditions were subjected to immunohistochemical analysis with an FITC-labeled antibody against α-smooth muscle actin (αSMA; Sigma, St. Louis, MO). Cells were plated at low density (10³ cells/mL) and analyzed 2 to 3 days after passage and at high density (10⁵ cells/mL) and analyzed 3 to 4 days after passage. At the specified time, cells were harvested and fixed (Histochoice; Amresco, Solon, OH) for 15 minutes, permeabilized with Triton-X-100, blocked with 2% BSA, and incubated with the FITC-labeled antibody. Control cells were treated in the same manner, with no antibody incubation. 

Electrophysiology

Cells were patch clamped using the amphotericin whole-cell perforated-patch technique. ¹⁶ Briefly, currents were recorded using a patch-clamp amplifier (model 200A; Axon Instruments, Burlingame, CA) and accompanying software (pClamp 8.0; Axon Instruments). Records were capacity compensated by the amplifier circuitry, sampled at 2 kHz, and filtered at 1 kHz. Cell capacitance was recorded for current density calculations (shown later). The pipette solution contained (in mM) 145 KOH, 100 methanesulfonic acid (MeS), 2.5 NaCl, 2.5 CaCl₂, 5 HEPES, and 240 mg/mL amphotericin B (Sigma). Unless otherwise noted, the bathing solution contained (in mM) 145 NaCl, 5 KCl (145 KCl, 2.5 NaCl for the KCl bath), 2.5 CaCl₂, 5 glucose, and 5 HEPES. To determine the permeability of ICl_LPA to MeS, an NaMeS-Cl bath solution containing (in mM) 65 NaCl, 65 NaMeS, 2.5 KCl, 1 CaCl₂, and 5 HEPES was used. Electrodes were coated (Sylgard; Dow Corning, Midland, MI) and fire polished. 

E _m was measured by noting the reversal potential (E _rev) from current-voltage (I-V) relationships. In instances in which currents were too small to reliably measure E _rev, or no current was active until the more depolarized voltage pulses, cells were current clamped to 0 pA, and E _m was read directly off of the patch-clamp amplifier digital display. Voltages were corrected for junction potentials by −15.4 mV, as estimated using the junction potential calculator included in the software package. The mobility of KMeS was set at 0.66 (Barry P, personal communication, 2002). In all cases, E _m comparisons were made before and after addition of agonists (and in some cases blockers) within the same cell (i.e., paired comparisons). 

ICl_LPA activation was examined under several conditions. In one series of experiments, different concentrations of LPA (100 nM–10 μM) were added to the bath to determine whether there was a dose-dependent effect of LPA on peak ICl_LPA current density (peak current/cell capacitance). In a second series of experiments, different concentrations of S1P (100 nM–10 μM) were added to the bath to determine whether S1P activates ICl_LPA and whether there was a dose-dependent effect of S1P on peak ICl_LPA current density. We also exposed the cells to a hyposmotic solution (220 mOsm) to confirm that the cultured cells contain the volume-activated Cl⁻ current seen in acutely isolated cells. Finally, permeability of MeS through ICl_LPA was determined by using the bath solution described earlier and a tail current protocol to measure E _rev. Cells were treated with 10 μM LPA, clamped according to an activation protocol (as in Fig. 2A ) to ensure that ICl_LPA was active, and then clamped with the tail-current protocol. For this protocol, cells were held at 0 mV, depolarized to 85 mV, and hyperpolarized to a series of voltages from 65 to −115 mV, in 15-mV intervals. The Cl:MeS permeability ratio was calculated with the E _rev measured when the cells were bathed in the NaCl and NaMeS-Cl baths, respectively, and with the Goldman-Hodgkin-Katz equation. ¹⁷  

Two additional studies were performed to confirm that ICl_LPA activation is receptor mediated. In the first experiment, the ability of a relatively inactive short-chain version of LPA (8:0 chain length; 10 μM) to activate ICl_LPA was compared with that of active LPA (18:1 chain length; 10 μM). In the second experiment, different concentrations of the LPA₁ and LPA₃ antagonist dioctyl-glycerol pyrophosphate (DGPP 8:0) were added to the bath with a constant LPA concentration (1 μM). DGPP has been shown to block LPA₃ and LPA₁ receptors with a K _i of 106 nM and 6.6 μM, respectively, and is ineffective at blocking LPA₂. ¹⁸ LPA₁ and LPA₃ are specific for LPA (S1P is not an agonist for these receptors). Thus, as a control, we also examined the effects of DGPP on S1P activation of ICl_LPA. A methanol control (DGPP is solubilized in methanol) was also examined. 

Chemicals

Unless otherwise stated, chemicals were obtained from Sigma. LPA (8:0 and 18:1), S1P, and DGPP were purchased from Avanti Polar Lipids (Alabaster, AL). 

Statistics

A paired Student’s t-test was used to compare results for all E _m studies. Analysis of variance (ANOVA) with least-squares means was used to compare the effects of different doses of LPA and S1P on the current density response. 

Results

Immunohistochemistry

Figure 1 shows the results of the phenotyping studies using the antibody against αSMA. Figure 1A shows positive αSMA staining in our standard cells (low density) used for the electrophysiological studies. Figure 1B shows relatively negative staining in high-density cells. Figure 1C shows the negative control (no antibody). 

Voltage-Gated Currents

Voltage-gated currents in the cultured keratocytes were similar to those we have previously reported for acutely isolated WAKs. More than 93% of cells examined (15/16) expressed the volume-activated Cl⁻ current (data not shown). Whereas a small number of cells had the delayed rectifying K⁺ current (Figs. 2A 2B 2D) and inward Na⁺ current (not shown) present in keratocytes acutely isolated from control rabbit corneas, ¹⁴ most had minimal expression of these currents (Figs. 2C 2D) , as is typical in acutely isolated WAKs. ³ The delayed rectifying K⁺ current and the inward Na⁺ current were seen in only 30% and 21% of cells examined, respectively. 

LPA-Activated Current

ICl_LPA was observed in most cells treated with 100 nM or more LPA (Figs. 3A 3B) . Average time to current activation after LPA exposure was approximately 5.6 minutes. This is similar to our finding in acutely isolated WAKs, ⁴ whereas the volume-activated current was active as soon as we began the protocol after a solution change in both cultured and isolated cells (within 15 seconds). As in cells acutely isolated from WAKs, the current began to inactivate at depolarized voltages and was blocked by 5-nitro-2-(3-phenylpropylamino) benzoic acid (NPPB; 50 μM; Fig. 3C ) and 4,4′-diisothiocyanostilbene-2,2′-disulfonic acid (DIDS; 100 μM: data not shown). Figure 3D shows the I-V relationship for the currents shown in Figures 3A 3B and 3C . ICl_LPA was activated in all cell types examined, whether or not they expressed the delayed rectifying K⁺ current. Figure 3E shows tail current I-V curves from a representative cell bathed in NaCl Ringer’s and in the MeS-Cl bath. As can be seen, E _rev in the NaCl and MeS-Cl solutions was −21 mV and −16 mV, respectively. If the current were carried by both Cl and MeS, the expected Nernst potentials would be −10 mV in the NaCl bath and −7 mV in the MeS-Cl bath. If the current were not carried by MeS, the expected Nernst potentials would be −107 mV and −86 mV in the NaCl and MeS-Cl bathing solutions, respectively. The mean (± SD) Cl-MeS permeability ratio equaled 1.22 ± 0.26 ( n = 5). Thus, the channel is almost as permeable to MeS as to Cl⁻. 

LPA increased the current density in a saturable dose-dependent manner (Fig. 4) . All concentrations of LPA except for 0.1 μM produced significantly elevated peak ICl_LPA current density compared with control (before LPA administration (0 μM; P ≤ 0.001). All concentrations of LPA were also significantly different from those in the the 1-μM and 10-μM groups ( P ≤ 0.001). In the DGPP experiments (Fig. 5) , 10 μM DGPP prevented ICl_LPA activation so that the current density of this group was significantly lower than the LPA control (1 μM; P = 0.001) and not different from the untreated control. Neither the lowest concentration of DGPP (1 μM) nor the methanol control significantly blocked the LPA-activated current. The volume-activated Cl⁻ current was not affected by DGPP (10 μM; n = 7; data not shown). 

S1P-Activated Current

S1P was also found to activate ICl_LPA (Fig. 6) . Currents had identical I-V relationships and blocker sensitivity as LPA-activated currents, with inactivation evident at depolarized voltages. Average time to current activation after S1P exposure was approximately 5.3 minutes, almost identical with that of the LPA-activated current. As with LPA, all groups produced a significantly higher peak current density than the control ( P < 0.001) except for the 0.1-μM group (Fig. 7) . Unlike LPA, which saturated at the 1-μM dose, 10 μM S1P produced a significantly greater peak current density than in the 1-μM group ( P < 0.005). As with LPA, the S1P-activated current was blocked by DIDS and NPPB (data not shown). DGPP (10 μM) was ineffective at blocking the 1 μM S1P-activated current ( P = 0.13). 

Resting Voltage

E _m was measured in cells before and after activation of ICl_LPA by both LPA and S1P and also after block of the LPA-activated current by DIDS and NPPB (Fig. 8) . Figure 8A shows that activation of ICl_LPA by both LPA ( n = 52) and S1P ( n = 12) resulted in significant cell depolarization ( P < 0.001). Figure 8B demonstrates that blockage of ICl_LPA by either DIDS ( n = 7) or NPPB ( n = 4) resulted in significant repolarization of the cells ( P < 0.001). Neither DIDS nor NPPB affected the unstimulated E _m ( n = 5 and 6, respectively; data not shown). 

Discussion

This study demonstrates that cultured rabbit keratocytes can serve as a useful model for studying changes in ion channels during wound-healing–induced shifts in cell phenotype and for studying PLGF activation of ICl_LPA. As in keratocytes freshly isolated from wounded rabbit corneas, rabbit keratocytes cultured in serum at low density had minimal expression of K⁺ and Na⁺ currents and had significant receptor-mediated activation of ICl_LPA. 

It has been shown that keratocytes grown in serum at low density become αSMA positive and thus are classified as myofibroblasts, which is the phenotype of keratocytes found in wound-healing conditions. ¹⁵ In contrast, cells grown under similar conditions at high density remain αSMA negative. In the presence of serum, these cells are typically of the fibroblast phenotype. ¹⁹ These high-density cells could be converted to αSMA positive in the presence of TGF-β. To establish and confirm the phenotype of the keratocytes used in the present study, immunohistochemistry was performed on cells cultured at low density (the cells we typically used for patch-clamp experiments) and at high density. These studies demonstrated that the cultured keratocytes we routinely use in the electrophysiological studies are αSMA positive, indicating that immunohistochemically they can be classified as cells of the myofibroblast phenotype. Because of the relative absence of K⁺ or Na⁺ currents and the ability of LPA to activate ICl_LPA in these cultured keratocytes, they can also functionally be classified as cells of the wound-healing phenotype. ³ Maltseva et al. ²⁰ have recently demonstrated that myofibroblasts can be reverted to the αSMA-negative phenotype. It will be interesting to see whether these reverted cells regain the K⁺ and Na⁺ currents lost in the initial transition to the myofibroblast phenotype. 

Previous studies examining the influence of LPA on WAK ion channels provided an initial characterization of the biophysical properties of ICl_LPA. They also provided some functional evidence that ICl_LPA is activated by LPA through a receptor-mediated response, as opposed to a cell volume–mediated response. ^{4 7} This evidence includes a dose-dependence study of LPA on peak amplitude of ICl_LPA current, failure of lysophosphatidyl choline (LPC) to activate ICl_LPA, and absence of activation of the current by LPA in a cell in which the current could be activated by an acute increase in volume. This last experiment demonstrated that the cell being examined was capable of expressing the Cl⁻ current, but in that instance, not through activation of the LPA receptor. The present study demonstrates that the dose-dependent effect of LPA on peak density of ICl_LPA current is also present in cultured keratocytes (Fig. 4) . The results demonstrating that the relatively inactive short-chain LPA (8:0) was significantly less effective at activating ICl_LPA than LPA (18:1) provide new evidence that LPA activates ICl_LPA through a receptor-mediated pathway. Because short-chain LPA is a much closer analogue to LPA than LPC, it serves as a more appropriate negative control for this type of study. That 10 μM short-chain LPA was not significantly different from the 100- or 300-nM LPA (18:1) groups in activating ICl_LPA indicates that the short-chain LPA either has a very weak affinity for LPA receptors or that it can possibly activate ICl_LPA through a yet to be described receptor or pathway. 

Additional evidence for receptor-mediated activation of ICl_LPA was provided by the experiments demonstrating inhibition of LPA activation of the current by the LPA receptor antagonist DGPP. In contrast, S1P activation of ICl_LPA was not inhibited by DGPP. Given that 1 μM DGPP was relatively ineffective at blocking ICl_LPA, whereas 10 μM was significantly effective, it appears that ICl_LPA activation is mediated through the LPA₁ receptor (DGPP blocks LPA₁ with a K _i of 6.6 μM versus LPA₃ with a K _i of 106 nM). ¹⁸ This must be confirmed in more detailed studies. 

LPA is known to be a strong activator of Rho. ²¹ Nilius et al. ²² provided strong evidence that the Rho-Rho kinase pathway is involved in the volume-regulated anion channel activation cascade in bovine vascular endothelial cells; thus, it is possible that LPA is acting through this pathway to activate ICl_LPA in WAKs. Postma et al. ²³ determined that LPA activation of an anion channel distinct from the volume-activated channel occurs through a Gα₁₃ mediated, Rho-independent pathway. It should be noted that activation of this Cl⁻ channel also depolarizes the cells. At this time, it is not possible to state which, if either, of these pathways is involved in LPA activation of ICl_LPA. 

To date, no function has been attributed to activation of ICl_LPA in corneal keratocytes. Because the LPA concentration surrounding and presumably within the cornea becomes elevated after corneal wounding, ⁵ it seems likely that there is significant ICl_LPA activity associated with this increased LPA concentration. This study clearly demonstrates that ICl_LPA activation results in cell depolarization and that block of ICl_LPA repolarizes the cell. Because this is the first study to measure E _m in any keratocyte model, it is not clear what the physiological consequences of the depolarization might be. It would be expected that the loss of the delayed rectifying K⁺ current in WAKs would lead to some degree of cell depolarization. ³ In our study ICl_LPA activation further depolarized the cells. Of course, significant depolarization alters the steady state ion concentrations and transport properties of the cell. This depolarization may be required for some of the activities performed by keratocytes during wound healing. In addition to the effects of depolarization on ion concentrations and transport, many studies have demonstrated that cell depolarization enhances the transcription of a number of different proteins. ^{24 25 26 27 28} This transcription response is typically (although not always) mediated by a depolarization-induced increase in intracellular Ca²⁺ concentration. Although there are no reports of depolarization-induced increases in keratocyte intracellular Ca²⁺ concentration, this possibility cannot be ruled out. Depolarization has also been shown to inhibit polyamine transport into ZR-75-1 human breast cancer cells. ²⁹ Polyamines have been shown to be critical for cell migration and mitosis and are involved in DNA transcription. ^{30 31} We have previously reported that disruption of polyamine synthesis in corneal cells, including keratocytes, inhibits proliferation. ³² Depolarization-induced inhibition of keratocyte polyamine transport could act to control keratocyte proliferation and migration during wound healing. 

This study also demonstrates that ICl_LPA is permeable to the relatively large anion, MeS. In many cell types, movement of large anions through volume-activated anion channels serves both a metabolic function and a regulatory function to decrease volume. These responses have yet to be studied in keratocytes. 

In conclusion, this work demonstrates that cultured corneal keratocytes can act as a useful model for the study of ion channel function in the WAK phenotype. As in keratocytes acutely isolated from wounded corneas, these cultured cells have little or no K⁺ or Na⁺ currents (the primary conductances in quiescent keratocytes), and they contain ICl_LPA (rarely seen in quiescent cells). This study provides detailed evidence that activation of ICl_LPA by LPA is receptor mediated and that ICl_LPA can also be activated by S1P. Finally, this study demonstrates that activation of ICl_LPA significantly depolarizes the cell. 

 Supported by National Eye Institute Grant EY12821 (MAW) and the University of Tennessee Center of Excellence for Diseases of Connective Tissue (LDC).

 Submitted for publication February 21, 2002; revised May 23, 2002; accepted June 3, 2002.

 Commercial relationships policy: N.

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 2.

 

Whole-cell currents from unstimulated cultured corneal keratocytes. (A, B, D) A small number of cells had a delayed rectifying K⁺ current. (C) Most cells had neither of these currents. (D) I-V relationship for the currents illustrated in (A–C). (○) Steady state I-V from current in (A); (•) steady state I-V from current from (B); (▵) steady state I-V from current from (C). Insets: voltage clamping protocol.

Figure 1.

 

αSMA phenotyping studies. (A) Positive αSMA staining in standard cells (plated at low density) used for electrophysiological studies. (B) Relatively negative staining in high-density cells. (C) Negative control (no antibody).

Figure 3.

 

LPA induced current (ICl_LPA) in a representative cultured corneal keratocyte. (A) Current from unstimulated cell. (B) Activation of ICl_LPA by 10 μm LPA. (C) Block of the current in (B) by NPPB (50 μM). (D) I-V relationship for the currents shown in (A–C). (○) Steady state I-V from current in (A); (•) steady state I-V (B); (▵) steady state I-V from current from (C). (E) Tail current I-V relationships from a representative cell bathed in NaCl Ringer’s (○) and NaCl-MeS solution (•).

Figure 4.

 

ICl_LPA current density after exposure to different concentrations of LPA or 10 μM short-chain (8:0) LPA. * P ≤ 0.001 versus control (0 μM group); † P ≤ 0.005 versus 1-μM and 10-μM groups. Experimental n listed in individual bars.

Figure 5.

 

Effect of DGPP on LPA-activated ICl_LPA. Note that 10 μM DGPP significantly inhibited the ability of 1 μM LPA to activate ICl_LPA. Neither the lowest concentration of DGPP (1 μM) nor the methanol control significantly blocked the current. †0 μM LPA group from Figure 3 ; * P < 0.001 versus control; Δ P ≤ 0.001 versus all groups except control. Experimental n listed in individual bars.

Figure 6.

 

S1P induced Cl⁻ current in a representative cultured corneal keratocyte. (A) Current from unstimulated cell. (B) Activation of Cl⁻ current by 10 μM S1P. (C) I-V relationship of the currents shown in (A) and (B). (○) Steady state I-V from current in (A); (•) steady state I-V from current in (B).

Figure 7.

 

Density of ICl_LPA after exposure to different concentrations of S1P. DGPP was ineffective at blocking the S1P-stimulated current. * P ≤ 0.001 versus control. Experimental n listed in individual bars.

Figure 8.

 

E _m in cells (A) before and after administration of LPA or S1P and (B) before and after block of ICl_LPA by DIDS or NPPB. All studies were performed in pairs. * P < 0.001 versus paired control.

The authors thank Gabor Tigyi for providing reconstituted, chemically active LPA, S1P, and DGPP and for helpful suggestions.