Synergistic Effects of a Calcium Ionophore and Activators of Protein Kinase C on Epithelial Paracellular Permeability Michael C. Winter, Michael W. Peterson, and D. Michael Shasby Department of Internal Medicine, University of Iowa College of Medicine and the Veterans Administration Hospital, Iowa City, Iowa

Oxidants reversibly increase the paracellular permeability of Madin Darby canine kidney (MDCK) epithelial cell monolayers, and the decrease in resistance occurs within 10 to 15 min of initiating oxidant exposure. Oxidants also initiate hydrolysis of phosphatidylinositol in MDCK cells, with resultant increases in diacylglycerol and inositol phosphates. Phorbol esters and synthetic diacylglycerols increase the paracellular permeability of MDCK monolayers with a time course similar to the oxidants. In contrast, calcium ionophores increase MDCK monolayer paracellular permeability only after 2 to 3 h of exposure. Because the products of the oxidant-initiated phospholipid hydrolysis would be likely to both activate protein kinase C and increase cell calcium, we asked if ionomycin, a calcium ionophore, and phorbol esters or diacylglycerols, activators of protein kinase C, might not act in concert to alter MDCK monolayer paracellular permeability. When ionomycin was added alone to MDCK monolayers, there was an increase in cell calcium, activation of a lumen negative current, a limited transitory decrease in transepithelial resistance, but no increase in mannitol flux across the monolayers. When phorbol dibutyrate (PDBU) or oleyl acetyl glycerol (OAG) were added to MDCK monolayers, there was no current activated, there was a progressive decrease in transepithelial resistance, and there was an increase in mannitol flux across the monolayers which was evident within 20 to 40 min of adding the agent. When 1 J.tM ionomycin was added to the monolayers along with PDBU or OAG, there was a synergistic increase in paracellular permeability of the monolayers when compared to addition of ionomycin, PDBU, or OAG alone. Inhibition of the lumen negative current with barium did not alter the increase in epithelial permeability. The synergistic effect of ionomycin and PDBU on paracellular permeability was paralleled by a synergistic increase in diacylglycerol formation in the cells. Agents that cause both an increase in cell calcium and activation of protein kinase C might be expected to have marked effects on epithelial paracellular permeability.

Control of the permeability of the paracellular pathway of various epithelia is an area of increasing interest. Pathologic changes in the permeability of this pathway occur during acute inflammation in several organs, and more recent evidence suggests that physiologic regulation of paracellular pathways may be important for normal transport across some epithelia (1-3). The quantitative and qualitative paracellular permeability of epithelia is generally felt to be primarily determined by the zonula occludens, or tight junction (4). Multiple observations have suggested important links between the tight junction itself and cytoskeletal structures, especially actin filaments (5, 6). Several investigators have reported that effects (Received in original form September 12, 1990 and in revisedform October 24, 1990) Address correspondence to: D. Michael Shasby, M.D., Department ofInternal Medicine, University of Iowa College of Medicine, Iowa City, IA 52242. Abbreviations: fetal bovine serum, PBS; Hanks' balanced salt solution, HBSS; Madin Darby canine kidney cell, MOCK cell; l-oieyl-z-acetyl-snglycerol, OAG; phorbol dibutyrate, PDBU. Am. J. Respir. Cell Mol. BioI. Vol. 4. pp. 470-477, 1991

of cytoskeletal active agents on cytoskeletal morphology and transepithelial permeability are dependent on ATP, suggesting the activity of a kinase in the control of epithelial permeability (7, 8). Both phorbol esters and diacylglycerols, activators of protein kinase C, increase paracellular permeability across cultured epithelial cell monolayers (9-12). In earlier reports, we and others found that oxidants reversibly increased the paracellular permeability of monolayers of Madin Darby canine kidney (MDCK) epithelial cells within 10 to 15 min of initiating exposure. We subsequently observed that oxidants initiated hydrolysis of MDCK cell inositol phospholipids with resultant increases in diacylglycerols and inositol phosphates (12-14). More recently, we found that the effects of phorbol esters and diacylglycerols on the paracellular permeability of MDCK monolayers were dependent on ATP, consistent with the idea that the effects of these compounds on paracellular permeability represent the activity of a kinase (15). While oxidants increased both diacylglycerol and inositol phosphates in MDCK cells, we also found that 1 J.tM ionomycin, by itself, did not increase MDCK paracellular permeability after 1 h (15). Peterson and Gruenhaupt (16) re-

Winter, Peterson, and Shasby: Ca-PKC Synergy and Epithelial Permeability

cently reported that 5-J.tM, but not 1-J,tM, solutions of the calcium ionophore, A23187, increased the paracellular permeability of MDCK cell monolayers. However, this occurred only after 2 to 3 h of exposure, a time course very different from the oxidant effects. Because oxidants increased both diacylglycerol, an activator of protein kinase C, and inositol phosphates, which release intracellular calcium, in MDCK cells, we wondered if the increase in calcium caused by 1 J,tM ionomycin, when combined with phorbol ester or diacylglycerol, would have an additive or synergistic effect on paracellular permeability that would occur within the time frame of the effects of oxidants.

Materials and Methods Materials Tissue culture media was obtained from the Cancer Center, University of Iowa. Fetal bovine serum (FBS) was from Hyclone (Logan, UT). Hanks' balanced salt solution (HBSS) was purchased from GIBCO (Grand Island, NY). Other salt solutions were made in the laboratory using the indicated concentrations of constituents. Polycarbonate micropore filters were from Nuclepore (Pleasanton, CA). A23187, Hepes, Pipes, dimethyl sulfoxide, EGTA, and EDTA were from Sigma Chemical Co. (St. Louis, MO). Fura-2 AM and Fura-2 were from Molecular Probes (Eugene, OR). (L4C]Mannitol and ")I-[32P]ATP were from New England Nuclear (Boston, MA). sn-1,2-diacylglycerol kinase from Escherichia coli was from Lipidex Inc. (Westfield, NJ). Cardiolipin C, sn-l,2dioctanoyl glycerol, and dioleoyl glycerol were from Avanti Polar Lipids (Pelham, AL). Cell Culture The MDCK cells used in this study were derived from a cloned line generously donated by Dr. Barry Gumbiner, University of California, San Francisco. They were cultured in minimal essential medium supplemented with 10% FBS, 2 mM glutamine, 100 U/rnl penicillin, and 100 }Lg/rnl streptomycin and buffered with 15 mM Hepes and 7.5 % sodium bicarbonate to pH 7.4. The cells were grown in tissue culture plates and passaged weekly in 10:1 split ratios using trypsin (0.25 %) and EDTA (1%). A O.S-ml suspension of 3 X 105 cells was placed on the filter in the lumen of a Millicell-HA culture plate (0.45 J,tm, 12 mm diameter) insert that fit snugly into a modified Ussing chamber (Jim's Instruments, Iowa City, IA). Monolayers were cultured in 24-well tissue culture plates and were ready for use 5 to 7 d after plating. Measurement of Monolayer Electrical Resistance and (l4C]Mannitol Flux The MDCK monolayers were mounted between the halves of the modified Ussing chambers (Jim's Instruments) (12). HBSS or the indicated buffer was used to bathe both surfaces. It was circulated by bubbling with 5 % CO 2 in air at 37° C. The electrical resistance was calculated from the change in current resulting from an intermittent transepithelial voltage pulse. Pulses of 3 mY, 2-s duration, with a 50-s period, were generated by a current-voltage clamp (Bioengineering, University of Iowa). In each instance, the baseline resistance was measured and observed to demon-

471

strate a stable pattern before initiating the experiment. Data are expressed as percentage of this baseline resistance. All monolayers had an initial resistance of > 1,000 ohm X em'. In most experiments, the flux of [14C]mannitol across the monolayers was also measured as an index of paracellular permeability. One millimolar of mannitol was added to both sides of the monolayer, and 5 J,tCi [14C]mannitol was added to the apical side. A baseline flux of labeled mannitol from apical to basal side of the monolayer was determined over two 20-min periods, the appropriate reagent was added to the indicated side of the monolayer, and the flux over two succeeding 20-min periods was measured. Data are presented as the mean of the two 20-min control periods and the two 20-min post-treatment periods. Measurements of Intracellular Calcium Epithelial cells on glass coverslips were loaded with 5 J,tM Fura-2 AM in the experimental buffer (135 mM NaCI, 1.2 mM CaCh, 1.2 mM MgCb, 10 mM Hepes, 10 mM dextrose, pH 7.4 with KOH) at 25° C X 60 min, as described (15). The cells were washed 5 times with the experimental buffer to remove free Fura-2, and the coverslips were then placed in a triangular cuvette in the experimental buffer. Fluorescence measurements were made in a Shimadzu model RF-540 spectrofluorometer using an excitation wavelength of 340 nm with a slitwidth of 2 nm and an emission wavelength of 510 nm with a slitwidth of 10 nm. Samples were scanned with an excitation wavelength of 300 to 390 nm to be certain of the appropriate fluorescence response for calcium-Fura-2. Maximal fluorescence was recorded after exposing the Fura-2-loaded cells to 0.1% Triton X-100, and minimum fluorescence was recorded subsequently by adding 10 mM EGTA (the Triton fluorescence was corrected for the decrease due to dilution of Fura-2 from the increased volume in the cuvette). Conversion of the fluorescence measurements to calcium concentrations was accomplished by measuring the fluorescence of known calcium concentrations in Fura-2 as described by Tsien (15). Analysis of Activation of Phospholipase C The amount of diacylglycerol produced in response to ionomycin, phorbol dibutyrate (PDBU), or both was determined by measuring the amount of 32P-Iabeled phosphatidic acid formed when diacylglycerol extracted from the cells

TABLE 1

Outline of experiments No Pretreatment

Barium

Dexamethasone

PDBU

OAG

PDBU

+

+

+

+

Ionomycin

Ionomycin

Ionomycin

Ionomycin

Transepithelial resistance and mannitol flux

Transepithelia I resistance and mannitol flux

PDBU

T ransepithelial resistance and mannitol flux

472

AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 4 1991

veloped in CHCkCH30H:CH3COOH (65:15:5 by volume). The plates were scanned with a Radiomatic Instruments model RS TLC Scanner. Specific activity of the ')I-(32P]ATP was determined by spotting dilutions onto a plate. The amount of 1,2-diacylglycerol in each sample was calculated using the specific activity and the known dilutions. Standards demonstrated greater than 80% conversion of diacylglycerol to phosphatidic acid. Statistical Analysis All data are reported as mean ± SE. Comparisons between groups were made either with a t test (two groups) or ANOVA, with differences between individual groups compared using a Tukey HSD test for post hoc comparisons of means. Groups that were different at the P ~ 0.05 level are described as different (including greater than or less than), and groups that were not different at this level of significance are described as not different. -1.0 +--.--.---.--.....-.----.-..,........,........-.--.......,---.-......-,--.-..............,

o

20

40

60

80

100

120

Time (min)

Figure 1. Dose-response effects of ionomycin on transepithelial resistance of Madin Darby canine kidney (MDCK) epithelial cell monolayers (n ~ 4 for each group). Ionomycin caused a transitory decrease in transepithelial resistance that persisted as long as the lumen negative current (see text).

was reacted with diacylglycerol kinase and ')I-[32P]ATP (17). Cells in 6-well plates were exposed to ionomycin, PDBU, or both as indicated. After 10 min, the medium was removed and the cells were immersed in ice-cold methanol and extracted according to the method of Bligh and Dyer (12) except that 1 M NaCI was used rather than water. The chloroform phase was washed twice with 1 M NaCI, and an aliquot was removed for measurement of lipid phosphorus. The chloroform phase was dried under nitrogen. Standards with 500 pmol dioleoyl glycerol were made at this point and treated the same as the samples for the rest of the procedure. The dried lipid was solubilized in 20 JLI of 7.5 % octyl-nglucoside and 5 mM cardiolipin C in 1 mM DETAPAC by sonication in an ultrasonic bath (50 to 60 Hz) for 15 s followed by brief vortexing and incubation at room temperature for 5 to 15 min. To the solubilized lipid/octyl-n-glucoside solution was added 50 JLI reaction buffer (l00 rnM imidazole HCI, pH 6.6, 100 rnM NaCI, 25 mM MgClz, and 2 mM EGTA), 2 JLI100 mM dithiothreitol, 5 JLI diacylglycerol kinase (stored at -20 0 C at 4 U/ml), and 13 JLI H20. The phosphorylation reaction was initiated by adding 10 JLI of 10 mM ')I-[32P]ATP (0.8 mCi/ml in 1 mM DETAPAC and 100 mM imidazole HCI, pH 6.6). The reaction proceeded for 30 min at 25 0 C and was stopped by extraction of the lipids as described for the cells. After separating the phases, the samples were stored overnight at -70 0 C. The chloroform phase was then washed twice with 2 ml 1% perchloric acid, and aO.5-ml aliquot was removed and dried under nitrogen. The sample was resuspended in 100 JLI of 5 % CH 30H in CHCI 3 • Twenty microliters of the resuspended sample was spotted onto 20 X 20 cm Silica Gel 60 A plates, which had been preactivated by running in acetone. The plates were then de-

Results Dose-response Effects of Ionomycin on Cell Calcium and Epithelial Permeability Addition of 1 JLM ionomycin increased MDCK cell calcium 228 ± 18 nM, and addition of 2 JLM ionomycin increased cell calcium 580 ± 200 nM (n ~ 3 for each). Addition of 5 JLM ionomycin increased the fluorescence signal as much as did the Triton, and it was therefore not possible to measure the increase in cell calcium.

POBU

0.2 CJ)

0

I:

as

--0--

lonomycin

----

POSU

----.--

---0---

lonomycin+POBU lonomycin+POBU

-0.0

U)

'(j) CD

c:c

-0.2

CD C'l I:

-0.4

.= as

J:

0

tU

I:

-0.6

0

:;:: 0

...

as u.

-0.8

-1.0 ~""""""'.,...-~-r-""""""""",,-'--""""""'~""""""'--'---'--' o 20 40 60 80 100 120

Time (min)

Figure 2. Transepithelial resistance of pairs of MDCK monolayers exposed to 1 110M ionomycin alone (open circles) versus 1 110M ionomycin with 0.1 110M phorbol dibutyrate (PDBU) (closed circles), or to 0.1 110M PDBU alone (open squares) versus 0.1 110M PDBU with 1 110M ionomycin (closed squares). The response was equivalent when ionomycin was added before PDBU (closed circles) or when PDBU was added before ionomycin (closed squares) (n ~ 5 for each group).

Winter, Peterson, and Shasby: Ca-PKC Synergy and Epithelial Permeability

473

OAG 0.2 G> 0

>< N

::J

E

:s '2

'E :

u::

c: ca

en

0

'2

'(j) G>

c: '0

.5

~

--0--

-0.0

a:

eE

--.--

-0.2

OAG 10nomycin+OAG

lonomycln

G>

C)

e ca

-0.4

.c 0

m c:

-0.6

0

;:: 0

... U.

ca -0.8

-1.0 0

50

100

150

Time (min)

Figure 4. Transepithelial resistance of pairs of MDCK mono layers . exposed to 20 J.'M 1-oleyl-2-acetyl-sn-glycerol (GAG) alone (squares) or 20 J.'M GAG with 1 J.'M ionomycin (triangles).

20

o POBU

POBU lono

Figure 3. Mannitol flux across pairs of MDCK monolayers during the control period and after treatment with 1 J.'Mionomycin alone versus 1 J.'M ionomycin with 0.1 J.'M PDBU (panel A) or 0.1 J.'M PDBU alone versus 0.1 J.'MPDBU with 1 J.'Mionomycin (panel B).

When pairs of MDCK monolayers were exposed to control medium or medium with 1, 2, or 5 ~M ionomycin, a lumen negative current was initiated with a transitory 10 to 20% decrease in transepithelial resistance as demonstrated in Figure 1. When mannitol fluxes across the same monolayers were compared for the control period versus the period after exposure to ionomycin, the change in flux for ionomycin-exposed mono layers was not different from that for control monolayers. Effects of Ionomycin with Phorbol Ester on Epithelial Permeability In contrast to the limited effects of ionomycin alone on epithelial permeability, when 1 p.M ionomycin was added along with 0.1 J.tM PDBU there was a rapid and persistent decrease in transepithelial resistance (Figure 2). The change in resistance in response to ionomycin and PDBU together was greater than that which followed ionomycin alone (1 J.tM), and both greater and earlier than that which followed PDBU alone (0.1 J.tM). The change in resistance with ionomycin and PDBU together was the same whether PDBU was added first or whether ionomycin was added first (the interval between additions was approximately 30 s), Paralleling the change in transepithelial resistance, the increase in mannitol flux across the same pairs of monolayers was greater when the monolayers were exposed to ionomycin and PDBU together than when they were exposed to each alone (Figures 3A and 3B). Similar to the results with PDBU, the change in transepithelial resistance was greater when 20 p.M 1-0Ieyl-2-

acetyl-sn-glycerol (GAG) was added to the monolayers with 1 p.M ionomycin than when the GAG was added alone (Figure 4). GAG with ionomycin also caused a greater change in mannitol flux (20.84 ± 3.23 pmol/min/cm-) than did GAG alone (4.04 ± 0.98 pmol/min/cm-). We next asked if ionomycin and PDBU together increased MDCK cell calcium more than ionomycin alone. Figures 5A through 5C are representative records from the fluorescence spectrophotometer of MDCK monolayers exposed to

A

EGTA

B

POBU

>-----~~----_/

~

Trit~n

lono

C

EGTA

~.~ POBU

~

lono

--------~-j~ I' Triton

I L

Figure 5. Representative tracings from fluorescent spectrophotometer measurements of Fura-2 fluorescence of MDCK cells loaded with Fura-2 and exposed to 1 J.'M ionomycin (panel A), 1 J.'M ionomycin and then 0.1 J.'MPDBU (panel B), or 0.1 J.'MPDBU and then 1 J.'M ionomycin (panel C). See text for details.

474

AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 4 1991

lonomycin + POBU

0.2 (l)

o

c: m

Control -0.0

Barium

tn

'0

(l)

II:

-0.2

.5 (l)

C)

e

-0.4

asc:

-0.6

m J: 0 0

:;:

o

...m

-0.8

LL

-1.0 0

50

100

150

Time (min)

c 'u>-

I ~:

~

E 0

c

.2

-

I

--

-

"t;

Figure 7. Change in transepithelial resistance of pairs of MDCK monolayers exposed to 1 JLM ionomycin and 0.1 JLM PDBU with (triangles) or without (squares) 10 mM BaCh added to the bath (n = 4 each).

1":'"""-

Figure 6. Representative tracings of current response to constant

voltage pulse and of short circuit current across MDCK monolayers exposed to 1 JLM ionomycin alone (panel A) or 1 JLM ionomycin and 0.1 JLM PDBU (panel B). Heavy black markings indicate time ionomycin and PDBU were added to the bath. The depression of the baseline represents short-circuit current. The height of the individual current responses to the voltage pulses is inversely proportional to the transepithelial resistance.

creased the lumen negative current activated in monolayers exposed to 1 p,Mionomycin and 0.1 JLM PDBU (2.24 ± 0.49 #-tA without barium and 1.14 ± 0.45 #-tA with barium; n = 4 each). However, addition of barium had no effect on the change in transepithelial resistance of monolayers exposed to 1 #-tM ionomycin and 0.1 #-tM PDBU (Figure 7), nor did it affect the change in mannitol flux across the same monolayers (Figure 8),

ionomycin alone, ionomycin and then PDBU, or PDBU and then ionomycin. Addition ofPDBU after ionomycin accelerated the rate at which cell calcium returned towards baseline (Figure 5B). Addition of PDBU before ionomycin reduced resting cell calcium a small amount (9 ± 1.7nM; n = 4) and markedly blunted the increase following ionomycin (calcium increased 228 ± 18.5 nM after 1 #-tM ionomycin alone and 43 ± 7.7 nM when 1 #-tM ionomycin was added after 0.1 #-tM PDBU; n = 4). Hence, the synergistic effect of ionomycin and PDBU together on transepithelial resistance was not due to an exaggerated increase in cell calcium. Role of Lumen Negative Current in Effects of Ionomycin and Phorbol Ester on Epithelial Permeability Addition of PDBU alone to MDCK monolayers did not cause activation of any current. However, addition of 1 #-tM ionomycin did activate a lumen negative current (3.64 ± 1.29 p,Apeak current). The peak current was not altered by addition of 0.1 #-tM PDBU along with 1 #-tM ionomycin (4.42 ± 0.31 #-tA peak current), but the duration of the current was prolonged with PDBU and ionomycin (49 ± 4 min) when compared to ionomycin alone (22 ± 4 min; Figure 6). Increased cell calcium activates basolateral potassium channels in MDCK cells. Addition of barium (10 mM), an inhibitor of potassium channels, to the bathing medium de-

80



Control Treated

60

>

Synergistic effects of a calcium ionophore and activators of protein kinase C on epithelial paracellular permeability.

Oxidants reversibly increase the paracellular permeability of Madin Darby canine kidney (MDCK) epithelial cell monolayers, and the decrease in resista...
811KB Sizes 0 Downloads 0 Views