点击显示 收起
【摘要】 Epithelia can adjust the permeability of their paracellular permeation route to physiological requirements, pathological conditions, and pharmacological challenges. This is reflected by a transepithelial electrical resistance (TER) ranging from a few tenth to several thousands ·cm 2, depending on the degree of sealing of the tight junction (TJ). The present work is part of an effort to understand the causes and mechanisms underlying these adaptations. We observed that an extract of human urine (hDLU) increases TER in a concentration- and time-dependent manner and is more effective when added from the basolateral side of cultured monolayers of Madin-Darby canine kidney cells than from the apical one. We found that its main TER-increasing component is epidermal growth factor (hEGF), as depletion of this peptide with specific antibodies, or inhibition of its receptor with PD153035, abolishes its effect. Since the permeability of the TJ depends on the expression of several species of membrane proteins, chiefly claudins, we explored whether hDLU can affect five members of the claudin family, the three known members of the ZO family, and occludin. EGF present in hDLU decreases the content of claudins-1 and -2 as well as delocalizes them from the TJ and increases the content of claudin-4. As expected from the fact that the degree of sealing of the TJ must be a physiologically regulated parameter, besides of hEGF, we also found that hDLU appears to contain also other components that decrease TER, claudin-4 and -7, and that seem to act with different kinetics than the TER-increasing ones.
【关键词】 claudin occludin transepithelial electrical resistance MadinDarby canine kidney monolayers kidney ZO
METAZOAN EXISTS THANKS TO transporting epithelia, at whose level organisms exchange substances with the environment. Since pioneering studies of this exchange were performed with tight epithelia like the frog skin and the urinary bladder, where the intercellular space (ICS) between cells is virtually hermetic, it was mainly understood on the basis of vectorial fluxes across the cells ( 12, 14, 19, 35 ). Later it was realized that when epithelia separate compartments contain isotonic fluids, such as the gall bladder and the ileum mucosa, the ICS is not tightly sealed, and the paracellular permeation route controlled by the tight junction (TJ) may play a paramount role ( 21 ). This focused the attention on TJ, that proved to be far more complex than initially suspected. Thus 1 ) the TJ was shown to be constituted by 50-odd protein species with a complex organization, that undergo phosphorylations/dephosphorylations and change their interrelationship between them, with the submembrane scaffold, as well as with the cytoskeleton in response to physiological and pathological conditions ( 10, 13 ). 2 ) Since the electrical resistance of the transcellular permeation route is generally much higher than the paracellular one, the overall electrical resistance of an epithelium (TER) reflects closely the degree of sealing of the TJ ( 9 ). 3 ) The TER of diverse epithelia varies over several orders of magnitude, in close correspondence with the difference in composition between the compartments separated by the epithelium. Thus the nephron has a TER 10 ·cm 2 in its proximal segment, which increases 10-fold in the distal, 100-fold in the collecting portion, and 1,000-fold in the urinary bladder. This increase is related to changes in structure ( 17, 18 ) and composition ( 49 ) of the TJ and is paralleled by an increase of concentration gradients between the interstitial fluid and the tubular content, which is gradually been converted into urine. In turn, this conversion to urine is due to 1 ) the operation of a multitude of translocating mechanisms present at the apical borders of the cells, such as ion channels, co- and countertransporters, aquaporins, etc. and 2 ) the paracellular route that allows the passage of large amount of isotonic fluid, chiefly in the proximal segment. The activity of these translocating mechanisms is under the control of a variety of hormones of the suprarenal gland, the parathyroid, the hypophysis, etc.
Since the wall of the proximal tube separates interstitial fluid from just filtrated plasma, but distal, collector, and vesical epithelia separate urine from interstitial fluid, the necessity of a strict correlation between content and degree of sealing of the TJ is teleologically obvious. Yet, we do not have a plausible explanation of this ability of the TJ to modulate its degree of sealing. To find a physiological cause, in a previous work we reasoned that if the fluid within the nephron contains a substance with the ability to increase TER in a concentration-dependent manner, and even reach urine without losing its activity, one would be able to find it. Accordingly, we prepared a lyophilized urine extract (DLU from D ialyzed and L yophilized U rine) that increases TER across monolayers of Madin-Darby canine kidney (MDCK) cells (epithelial from dog kidney) ( 26 ). In the present work, we pursue the analysis of human DLU by studying one of the most prominent proteins present in urine, the epidermal growth factor (EGF) ( 20, 32 ). We show that EGF modifies TER by changing the amount and distribution pattern of different claudins, which are known to be membrane components of the TJ, involved in permeation through the paracellular route ( 25, 58, 60, 61 ).
Urine composition is not constant, but it varies according to a multitude of physiological variables, among them the degree of hydration of the body, and the secretion of hormones from the hypophysis, the parathyroid and adrenal glands. The final composition is adjusted primarily in the distal segments. On this basis, it is expected that urine will not merely contain TER-increasing factors but TER decreasing as well. Accordingly, besides EGF we also find a TER-decreasing component that acts with somewhat slower kinetics.
MATERIALS AND METHODS
Cell culture and chemicals. Starter MDCK cultures were obtained from the American Type Culture Collection (MDCK, CCL-34). Upon arrival, cells were cloned and all experiments were performed in cells of subclone 7.11, chosen because of its intense blistering activity when plated on impermeable supports and a low basal TER. Cells were grown at 36.5°C in a 5% CO 2 atmosphere. DMEM (GIBCO-Invitrogen, Carlsbad, CA), with penicillin-streptomycin (10,000 U·µg -1 ·ml -1, In Vitro, Acayucan, Mexico, DF), and 10% donor bovine serum with iron (GIBCO-Invitrogen) were used. This medium will be referred to as CDMEM. Cells were harvested with trypsin-EDTA and plated on inserts (3.0-µm pore size, 6.5-mm diameter, 0.33-cm 2 Transwell, Corning Costar, Cambridge, MA) at 2.25 x 10 5 cell/cm 2, maintained 55 h in CDMEM, followed by serum starvation (24 h in DMEM) and then treated with the different media.
Human epidermal growth factor (hEGF) was purchased from Sigma (St. Louis, MO). The inhibitor of the EGF receptor (EGFR) PD153035 was purchased from Calbiochem (EMD Biosciences, La Jolla, CA). hEGF neutralizing antibody was purchased from Abcam (Cambridge, MA). Rabbit anti-claudin-1, claudin-2, claudin-3, claudin-7, occludin, ZO-1, ZO-2, and ZO-3, mouse anti-claudin-4, and horseradish peroxidase (HRP)-goat anti-mouse, FITC-goat anti-rabbit, and anti-mouse were purchased from Zymed (South San Francisco, CA). HRP-donkey anti-rabbit was purchased from Promega (Madison, WI). Mouse anti-actin was a kind gift of Dr. J. M. Hernández (Department of Cell Biology, CINVESTAV) ( 31 ).
hDLU preparation. Human urine was collected from 40 clinical informed healthy males between 18 and 23 years old that accepted had not taken any drug for at least a week before the collection. All the samples were from the first urine of the day and were collected the same day. Each sample was tested in a urine exam and after discarding the samples that had not the normal values, all the remaining samples were mixed and frozen immediately. The urine collected was centrifuged 30 min at 17,000 g and 4°C. Then, urine was extensively dialyzed against deionized water (100 ml of urine against 16 l of water distributed in 8 washes) in a 7,000 MW cutoff dialysis tubing (SnakeSkin Pierce, Rockford, IL). It was then lyophilized in a Labconco Lyph Lock 4.5 (Kansas City, MO) apparatus. The powder obtained was reconstituted in deionized water and the content of total protein was determined using Bradford?s method. By this procedure the human urine is concentrated 20 times. The extract obtained in this way (hDLU for h uman d ialyzed and l yophilized u rine) was maintained at -20°C, defrosted the day of the experiment, and diluted in DMEM or CDMEM. The employed protocol was approved by the Ethics Committee for Human Studies CINVESTAV (IRB0004785-CINVESTAV IRB#1-COBISH).
TER. The degree of sealing of TJ to ionic solutes was assessed by measuring the TER using an EVOM and an EndOhm-6 systems (World Precision Instruments, Sarasota, FL). The cells grown on Transwell permeable supports were equilibrated at room temperature during 10 min. Final values were obtained by subtracting resistance of the bathing solution and the empty insert, and the results are expressed as ohms·cm 2 ( ·cm 2 ).
hEGF depletion. hDLU was incubated with the hEGF neutralizing antibody (20 µg of antibody per 1 ml of hDLU) overnight at 4°C with gentle agitation. Then, protein G agarose beads (Upstate, Lake Placid, NY) were added and incubated for 1 h and precipitated by centrifugation at 9,000 g for 1 min. The supernatant the hDLU not treated with the anti-EGF antibody was diluted in DMEM at the same proportions.
Paracellular permeability. Cells were grown on Transwell supports as described above, treated for 16 h with hDLU (5 µg/ml) or hEGF (100 ng/ml), and after TER measurement, the monolayers were washed twice with P buffer (10 mM HEPES, pH 7.4, 1 mM sodium pyruvate, 10 mM glucose, 3 mM CaCl 2, 145 mM NaCl). Then, monolayers were incubated for 30 min with hEGF (100 ng/ml) or hDLU (5 µg/ml) diluted in P buffer and, after that, the apical medium was replaced with a solution containing 10 µg/ml of FITC-dextran of either 3 or 20 kDa. After 1-h incubation at 37°C, the basal medium was collected and the fluorescence of the transported FITC-dextran was measured with a fluorescence spectrometer LS-3B (Perkin-Elmer) at ex 492 nm and em 520 nm.
Western blot. After TER measure, monolayers in inserts were washed three times with ice-cold TRIS-buffered saline solution (TBS: TRIS 100 mM, pH 7.5, 100 mM NaCl) and then incubated at 4°C for 20 min with buffer lysis (10 mM Tris, pH 7.5, 1 mM sodium orthovanadate, 1% SDS) for protein extraction. The extracts were passed 10 times through an insulin syringe and centrifuged 20 min at 17,000 g. The supernatant was recovered, its total protein content measured by the BCA assay (Pierce), subsequently boiled in Laemmli sample buffer (50% glycerol, 125 mM Tris·HCl, pH 6.8, 4% SDS, 0.125% bromophenol blue, 5% -mercaptoethanol) and resolved by SDS-polyacrylamide gel electrophoresis and transferred to PVDF sheets (Hybond-P, Amersham Biosciences, Piscataway, NJ). These sheets were blocked overnight with 3% BSA, and proteins of interest were detected with the specific antibody, followed by species-appropriate peroxidase-conjugated antibodies and a chemiluminescent system for detection (ECL Amersham Biosciences). The resolved bands were then analyzed with the software Kodak 1D 3.5.4 (Eastman Kodak, Rochester, NY) and the data were processed using GraphPad Prism 3 (GraphPad Software, San Diego, CA). The densitometric values were normalized with the signal obtained for actin, because we observed that the content of this protein does not change significantly in the different experiments. Once blotted for one protein, the membranes were stripped with 2% sodium dithionite in TBS with 0.4% Tween and reproved for another one. This treatment does not alter the signal obtained for the first protein blotted, since if it is reblotted and analyzed after at least six strippings, the result is the same as in the first blot (not shown).
Immunofluorescence. After TER measure, monolayers in inserts were washed three times with ice-cold TBS, fixed, and permeabilized methanol or ethanol/acetic acid (95%/5% vol/vol) for 5 min at room temperature, washed with TBS, blocked for 30 min with 3% BSA, and treated for 1 h at 37°C with a specific first antibody indicated above. Monolayers were then rinsed three times with TBS, incubated with a FITC-labeled suitable antibody for 30 min at room temperature, rinsed as indicated before, mounted in Fluorguard (Bio-Rad), and examined by confocal microscopy SP2 (Leica Microsystems Wetzlar, Wetzlar, Germany). Images captured were imported into Adobe PhotoShop (Adobe Systems, Mountain View, CA).
Statistical analyses. Statistical analyses were performed with GraphPad Prism 3. Results are expressed as means ± SE. Statistical significance in an unpaired two-tailed t -test is indicated as follows: * P < 0.05, ** P < 0.005, *** P < 0.001; n = number of observations. For correlation tests, a Pearson correlation was used and the coefficient of determination ( r 2 ) is reported with its correspondent P value.
RESULTS
The value of TER in monolayers of MDCK cells exposed to hDLU increases with time, reaching a maximum in 16 h ( Fig. 1 A ). This effect can be entirely ascribed to changes in the TJ on two grounds. First, the electrical resistance of the plasma membrane of MDCK cells is so high, the transcellular route could not account for the effects observed in the present work. Second, Drs. A. Ponce and J. J. Gallardo, using whole cell clamp techniques, have shown that dDLU (notice: DLU from dog, not from humans as in the present work) does not vary membrane resistance (unpublished results). Interestingly, TER decreases thereafter to a value below control, suggesting that it might not be due just to an inactivation of the TER-increasing factor, nor to a desensitization of its effector, but to a TER-decreasing component that would either become evident upon saturation of the first one or act with much slower kinetics. Figure 1 B shows the maximum value achieved at the 16th h as a function of the concentration of hDLU. On the basis of these results, we adopted a concentration of 5 µg/ml for 16 h as a standard condition to perform the assays reported below. We also observe that TER increases with the addition of hDLU to either side of the monolayer ( Fig. 2, column 4 ), albeit the basolateral response (53%) is significantly higher than the one of the apical side (7%). This would suggest that there are more receptors on the basolateral domain or that those in this position have a higher sensitivity.
Fig. 1. hDLU enhances the transepithelial electrical resistance (TER) of Madin-Darby canine kidney (MDCK) cells in a concentration- and time-dependent manner. A : concentration refers to the quantity of proteins added by dilution of hDLU in DMEM. hDLU enhances TER by the 8 h and reaches a maximum effect at 16 h with a hDLU of 5 µg/ml; n = 3 for each condition. B : maximum TER achieved at 16 h as a function of the concentration of hDLU.
Fig. 2. hDLU acts primarily from the basolateral side. In this and following figures, shaded columns refer to monolayers treated with hDLU for 16 h with 5 µg/ml hDLU. In this experiment, the extract was added to the basolateral side, the apical, and to both sides, respectively. Notice that when added to both sides the effect of hDLU is not equal to the sum of the independent effects on the basolateral plus the apical treatment; n = 3 for each condition. ** P < 0.005, *** P < 0.001.
Gallardo et al. ( 26 ) showed that the TER-increasing effect of dog DLU is due to a component that can be inactivated by proteases but not by heating 10 min at 92°C. On this basis, we investigated the role of hEGF, because it is a prominent protein of urine ( 20, 32 ), is heat stable ( 33 ), interacts with receptors present in epithelia, and these receptors are located on the basolateral side ( 24 ). Figure 3 shows that if the hEGF present in urine is neutralized by the presence of a specific antibody, the increase of TER is canceled (second vs. third column). By itself, the antibody does not modify TER ( column 4 ). An isotype control antibody by itself cannot modify TER either ( column 5 ), and when this isotype control antibody is added together with hDLU it cannot prevent the TER-increasing effect of the extract ( column 7 ). When hDLU is depleted beforehand with anti-hEGF-specific antibody ( column 7 ), the extract is no longer able to increase TER. These results indicate that the TER-increasing factor in hDLU is hEGF and suggest the existence of factors that decrement TER as well, because hDLU slightly but consistently diminishes TER when hEGF is neutralized (third and seventh columns).
Fig. 3. Neutralization of hEGF inhibits TER enhancement by hDLU. Monolayers were treated with hDLU alone ( column 2 ) or supplemented with an antibody against hEGF (1 µg/ml, column 3 ). By itself, the antibody, i.e., without hDLU, does not perturb TER ( column 4 ). An isotype control antibody does not perturb TER either ( column 5 ) nor prevents the TER-increasing effect of hDLU ( column 6 ). While in these cases hDLU was added simultaneously with the respective antibodies, in the last column hDLU was depleted of hEGF by immunoprecipitation before the exposure of the monolayer; n = 3 for each condition.
To investigate whether hEGF alone is sufficient to enhance TER, we next evaluated the effect of recombinant purified hEGF on the TER of MDCK monolayers. Figure 4 shows that hEGF at concentrations ranging from 50 to 200 ng/ml increment significantly the TER by 8 h and reaches a maximum effect at 16 h. Interestingly, the increase in TER induced by 200 ng/ml hEGF ( Fig. 4, dashed line) is lower than the ones produced by 100 ng/ml hEGF or 5 µg/ml hDLU. Thus hEGF alone is able to enhance the TER and does not seem to require other factors present in urine.
Fig. 4. hEGF enhances the TER in a concentration- and time-dependent manner. The effect of different hEGF concentrations on the TER of MDCK monolayers at the indicated times is shown. hEGF is found to enhance TER by the 8 h, and reach a maximum at 16 h, with concentrations ranging from 50 to 200 ng/ml. Absence of errors indicates that these are smaller than symbols. After 16 h TER begins to decrease, until it reaches control values at 24 h (not shown). It may be noticed that the curve corresponding to 200 ng/ml (dashed line) is always below both, the one corresponding to 100 ng/ml, and the one recorded with hDLU; n = 3 for each condition.
TJ also regulates the movement of nonionic molecules through the paracellular pathway. Therefore, we studied whether hDLU or hEGF modulates the paracellular permeability of MDCK monolayer to both 3 and 20 kDa FITC-dextrans. Figure 5 A shows the usual increment of TER of MDCK monolayers induced by hDLU and hEGF at the moment that were used to measure the apical to basolateral flux of FITC-dextrans. hDLU does not modify significantly the flux of either tracer ( Fig. 5 B, gray bars). On the other hand, treatment with hEGF ( Fig. 5 B, black bars) significantly augments the paracellular flux of the 3 kDa FITC-dextran but, interestingly, tends to diminish that of the 20 kDa ( P = 0.0538). As shown in Fig. 5 C, the TER and the paracellular permeability for the 3 kDa FITC-dextran changes are independent (hDLU: r 2 = 0.2226, P = 0.1998; hEGF: r 2 = 0.2991, P = 0.0657).
Fig. 5. Paracellular permeability to both ionic and nonionic molecules is independently modulated by hDLU and hEGF. TER ( A ) and paracelluar permeability ( B ) to 3 and 20 kDa FITC-dextran of MDCK monolayers treated for 16 h with 5 µg/ml hDLU (gray bars) or 100 ng/ml hEGF (black bars). * P < 0.05, P = 0.0538. C : correlation between TER and paracellular permeability to 3 kDa FITC-dextran of MDCK monolayers treated with hDLU ( r 2 = 0.2226, P = 0.1998) or hEGF ( r 2 = 0.2991, P = 0.0657).
EGF binds to and activates specific receptors on the cell membrane by triggering different signaling cascades that modify cell behavior ( 7, 8 ). We found that when the EGF receptor (EGFR) is inhibited by the use of PD153035 ( Fig. 6 ), that is one of the most potent and specific EGFR inhibitors available ( 23 ), the increment in TER induced by hEGF ( Fig. 6, second column) is reduced (fourth column) and that of hDLU (fifth column) is completely abolished (sixth column).
Fig. 6. Inhibition of epidermal growth factor receptor (EGFR) blocks the increment of TER induced by both hDLU and hEGF. TER of MDCK monolayers treated for 16 h with 100 ng/ml hEGF ( column 2 ) or 5 µg/ml hDLU ( column 5 ), in the absence ( column 3 ) or presence of a specific EGFR inhibitor (2.5 µM PD153035; columns 4 and 6 ). The inhibition of EGFR effectively blocks the increment of TER induced by hDLU and significantly reduces the effect of hEGF. Means ± SE, n = 10 for each condition.
In our previous work ( 26 ), we showed that the TER increment induced by dog DLU is not accompanied in changes in neither the number of TJ strands or its complexity. This does not preclude that the TJ would change by expressing other combinations of occludin and claudins, which are known to comprise families whose members have different permeability properties. We investigated whether hDLU and hEGF provoke changes in the cellular distribution of claudins (cln)-1, -2, -3, -4, -7, occludin, and the membrane-associated proteins zonula ocludens (ZO)-1, -2, and -3 ( Fig. 7 ).
Fig. 7. hDLU and hEGF modify the localization of specific tight junction (TJ) proteins. Representative images of localization of claudins (cln)-1, -2, -3, -4, -7, occludin, ZO-1, and ZO-3 in MDCK cells treated with hDLU (5 µg/ml) or hEGF (100 ng/ml) for 16 h. Monolayers can be seen in front and transverse optical sections. Scale bar = 10 µm. The panels below each en face image show xz section of the corresponding panels. The signals for cln-1 and -2 are reduced by both hEGF and hDLU. hDLU severely affects the distribution of cln-7. On the other hand, hEGF enhances the localization of both cln-4 and -7 on the lateral domain of MDCK cells. No other protein was altered in an evident way.
Treatment of MDCK monolayers with 5 µg/ml hDLU or 100 ng/ml hEGF induce a reduction of cln-1 staining, which nevertheless remains associated to cell-cell regions, as observed by confocal microscopy. hDLU and, to a greater extent, hEGF also perturb the localization of cln-2 and diminish the intensity of stain, which correlates with the increment in TER observed, since cln-2 enhance cation paracelluar permeability ( 25 ). Interestingly, in MDCK cells treated with hDLU and hEGF, the distribution of cln-7 is modified but in different ways. hDLU stimulates both a reduction in the number of cells stained and a disorganization of the cln-7 pattern at some sites (arrows); however, there are also increments in staining at some cell-cell borders (arrowheads). On the contrary, the effect of hEGF on cln-7 is homogenous, causing an increment of the protein localized all along the lateral membrane as observed in the xz section. This last effect is also observed for cln-4 in cells treated with hEGF, but not in those treated with hDLU. Occludin, ZO-1, and ZO-3 do not seem to be changed by these treatments. Comparison of the effects of hDLU and hEGF on cln-1, cln-4, and cln-7 indicates that hEGF present in urine does not exactly reproduce the effect of hEGF alone and suggests that urine contains other factors that increase TJ permeability possibly through modifications of cln-7. The changes in different TJ proteins, which can be induced with hDLU and hEGF, may be in principle due to changes in the quantity of these proteins, or else to changes in their associations with other proteins.
Figures 8 and 9 show a representative blot and the densitometric analysis, respectively, using actin as a load control, as this protein does not change under the conditions tested. hEGF and hDLU decrease the cell content of cln-1 by 50%, an effect that can be blocked by EGFR inhibitor PD153035. Levels of cln-2 decrease 75% with hEGF and hDLU. Interestingly, inhibition of the EGFR by itself increments the quantity of cln-2 by 200%, independently of hEGF or hDLU, suggesting that EGFR basal signaling exerts a negative control of cln-2 expression. The cell content of Cln-3 is significantly augmented only by PD153035 treatment (not shown). The level of cln-4 was enhanced to a similar extent by both hEGF and hDLU ( 50%) and PD153035 inhibited the effect of hEGF. However, when EGFR is inhibited and the monolayers are treated with hDLU there is a significant decrement of cln-4. Finally, levels of cln-7 are reduced significantly by hDLU treatment, independently of the inhibition of EGFR, whereas hEGF treatment does not have an appreciable effect. These results show that hEGF, either already present in hDLU or added experimentally, increments TER by specifically varying TJ proteins. Furthermore, these results also strongly suggests that some factors as yet not identified in human urine elicit on the TJ some effects that differ from that of hEGF, such as reducing TER by altering the composition of at least cln-4 and -7.
Fig. 8. hDLU and hEGF change the cell content of specific TJ proteins. Representative images of Western blots of claudins (Cln)-1, -2, -3, -4, -7, and actin in MDCK cells treated for 16 h with 5 µg/ml hDLU, or 100 ng/ml hEGF, in the absence or presence of a specific EGFR inhibitor, PD153035 (2.5 µM).
Fig. 9. Effects of the inhibition of the EGFR on the content of several claudins. Densitometric analysis of the bands resolved in Western blots. To ease comparison between different claudin species, the value for a given claudin was divided by its own control; n = 7 for each condition (except PD153035, n = 6). * P < 0.05, ** P < 0.005, *** P < 0.001.
DISCUSSION
The degree of TJ sealing varies over five orders of magnitude in different epithelia, from those of leaky proximal tubule, to a virtually hermetic colon and urinary bladder. Tightness can also change in the same epithelium according to physiological or pathological conditions, in response to pharmacological challenges ( 2, 28 ), or increase in a remarkable progression along the nephron as the glomerular filtrate is gradually converted into urine ( 11, 13, 14, 54 ). The molecular counterpart of this physiological versatility of the TJ is constituted by 50-odd protein species, whose physiological roles await elucidation. The present work is part of an effort to find molecules and mechanisms that may account for this variability.
In a previous work ( 26 ), we found that extracts of dDLU contain at least one protein species that is heat stable and increases TER of MDCK monolayers. Since Gallardo et al. ( 26 ) found that urine extracts from different animal species, such as cat, rabbit, and human, also increment the TER of MDCK monolayers, in the present work, we now pursue the study with a similar extract of hDLU. We found that hDLU, like dDLU, increments the TER of MDCK monolayers in a concentration- and time-dependent manner with a stronger effect from the basolateral than from the apical side, and the simultaneous addition to both sides does not elicits a greater response. Yet, it may be premature to speculate on sidedness at this moment. We also show that part the TER-increasing activity of hDLU is due to its content of hEGF, since the effect of the extract can be blocked by neutralization or immunoprecipitation of hEGF with a specific antibody, as well as by inhibiting EGFR with PD153035.
Among the molecular components of the TJ that may underlie the effect of hDLU, there are some, such as occludin and the claudin family, that pertain to the plasma membrane and are known to influence both electrical conductance and permeability to nonionic extracellular substances such as mannitol and dextran ( 3, 25, 58, 63 ). Treatment of MDCK monolayers with 5 µg/ml hDLU or 100 ng/ml hEGF reduce cln-1 staining, which nevertheless remains associated to cell-cell regions. hDLU and, to a greater extent, hEGF also perturb the localization of cln-2 and diminish the intensity of its staining in immunofluorescent studies, which correlates with the increment in TER observed, since this claudin enhances cation paracelluar permeability ( 25 ). hDLU stimulates a reduction in the number of cells stained for cln-7, a disorganization of the pattern at some groups of cells ( Fig. 6 ), and a decrease in its cell content ( Figs. 7 and 8 ). On the contrary, the effect of hEGF on cln-7 is homogenous, causing an increment of the protein localized on the lateral membrane ( Fig. 6, xz sections). This last effect is also observed with cln-4 in cells treated with hEGF, but not in those treated with hDLU. Furthermore, when the EGFR is inhibited, hDLU decreases the content of cln-4 instead. Other TJ components, such as ZO-1, ZO-2, ZO-3, and occludin do not seem to be changed by these treatments. Interestingly, it has been suggested that, in addition to its role as regulators of TJ permeability, different claudins might be involved in cell-cell adhesion, and several members of this family have been located along the basolateral domain in different epithelia regardless of its localization at the TJ ( 1, 30, 36, 37 ). Thus the changes in the localization of cln-4 and -7 possibly involve the interaction of these proteins with other cell adhesion molecules, among them the epithelial cell adhesion molecule EpCAM ( 37 ).
We also make several observations suggesting that, besides of its TER-increasing component, hDLU would be eventually found to contain TER-decreasing molecular species: 1 ) once hDLU has reached a peak of TER, this parameter decreases with longer exposure times ( Fig. 1 ). 2 ) At concentrations of 5 µg/ml, the response to hDLU is higher than the one at 10, 20, and 30 µg/ml ( Fig. 1 B ). 3 ) When hEGF is neutralized with specific antibodies, TER is slightly but significantly decreased ( Fig. 3 A ). Comparison of the effects of hDLU and hEGF on 4 ) paracellular permeability to nonionic molecules and on 5 ) cln-4 and cln-7 indicates that hEGF present in urine does not exactly reproduce the effect of hEGF alone ( Figs. 5, 7, and 8 ). Taken together, this would suggest that the putative TER-decreasing component is either slower or becomes evident once the TER-increasing component has reached saturation. Furthermore, physiological regulations are seldom, if ever, due to a single factor displacing a parameter in one direction, justifying the speculation that hDLU may contain agents that enhance as well as others that reduce TER. Moreover, the fact that the effects take hours before achieving a peak may be accounted for elaborated structural changes.
EGF has been previously shown to increment the TER of canine oxyntic mucosal cells ( 15 ), LLC-PK 1 cells ( 51, 57 ), alveolar epithelial cells (AEC) ( 5 ), and MDCK-II cells ( 55 ). In AEC, EGF stimulates the expression of cln-4 and -7 and downregulates cln-3 and -5 ( 16 ). Singh and Harris ( 55 ) reported that EGF treatment of MDCK-II cells increments both the expression and localization to the TJ of cln-1, -3, and -4, whereas downregulates the expression of cln-2. The differences between our results and those of Singh and Harris may be due to differences in the cell clones used.
EGF is synthesized in the thick ascending limb of Henle and distal convoluted tubule where it is localized at the apical membrane ( 27, 52 ). In human urine, EGF reaches a concentration of 50 nM ( 56 ) of which 50% is present in high molecular weight forms (165, 116, 97, 66, 50, 42, and 30 kDa) and the rest appears as the mature form (6 kDa) ( 38, 43, 59 ). This suggests that EGF liberated to the lumen is processed through its way along the nephron. Mature EGF corresponds to the 5% of its precursor, opening the possibility that the cleaved part of the peptide would have regulatory functions. In fact, immature forms of EGF are less active than the mature form ( 44 ). We propose that the absorption of water builds up a progressive concentration gradient of along the nephron. Thus, at more distal segments EGF would become more active by virtue of being more concentrated or processed, than at the sites of the nephron where is synthesized and delivered to the lumen. In this respect, it has been shown that in suckling as well as adult rats, there exists a proximodistal gradient of EGF content along the small intestine ( 53 ). In addition, extracellular regulated kinases 1/2 (ERK1/2), effectors involved in the signal transduction of EGFR, have been shown to downregulate the expression of cln-2 in renal epithelial cells ( 39 ), and active ERK1/2 is largely restricted to the cytoplasm of the distal nephron segments in normal human kidney ( 40 ), where cln-2 is absent ( 49 ). Thus EGF (or other growth factors of this family) may play a role in maintaining the progressive sealing of the TJ in organs that need to concentrate and maintain the progressive sealing of the TJ in organs that need to concentrate and maintain the composition of its fluid content, such as the kidney, the gastrointestinal tract, and the mammary gland.
Furthermore, it has been reported that ErbB2 is apically expressed along the nephron in human fetal kidneys ( 6 ). In these early stages of development, the concentrating function has not been yet achieved and EGF expression is low or undetectable. In human, urinary EGF excretion increases sixfold from birth to the second year of life and then decreases progressively until late senescence ( 41, 42 ), and in the mouse an 100-fold rise of the urinary EGF concentration occurs between birth and maturity ( 48 ). Thus, on the onset of EGF expression, the apical receptors may be responsive and EGF might be involved in the establishment of the progressive sealing of the TJ. Moreover, apical low affinity binding sites for EGF have been detected in mature rat kidney ( 34 ) and may also participate in the suggested mechanism.
On the other hand, when this system has reached maturation, the EGF and other members of this family of growth factors present in the fluid that bathe epithelia (urine, milk, saliva, bile, seminal plasma, tears, airway surface liquid) at concentrations several orders higher than that found in plasma ( 4, 22, 32, 46, 47, 50 ) may participate in an efficient repairing system. Since receptors are located on the basolateral side, they might not be accessible to its ligands. The physical segregation of a ligand to the apical solution and its receptor to the basolateral membrane provides a simple system that is primed for activation the instant epithelial integrity is compromised. Such a system could play a major role in the rapid restoration of barrier function and protection of the organism following an insult that disrupts the epithelium, as has been demonstrated in airway epithelia in vitro ( 62 ).
In recent years, opening of the TJ by extracellular substances has been studied as a strategy to improve drug absorption through epithelia ( 29, 45 ). One of the limitations of this strategy is reverse the effect and seal the TJ once the agent penetrates. Our findings open the possibility to use EGF (or similar TER-increasing factors) to enhance the sealing of the TJ after the treatment with agents that open them.
The ocean acts as a constant reservoir with respect to unicellular organisms. On the contrary, cells in a metazoan exchange substances with a milieu which has a smaller size than the cellular compartment and that would be quickly spoiled were it not for a circulatory apparatus that continuously carries fluids toward and away from epithelia with a comparatively large area (intestine, kidneys, lungs, gills) and whose cells specialize in vectorial and highly specific exchange. Moreover, higher organisms have hormones such as those of the hypophysis, parathyroid and the suprarrenal glands, that modulate this exchange through the transcellular route. Therefore, it does not come as a surprise that the paracellular route would be regulated as well. Actually, the apical side of the nephron is exposed to a glomerular filtrate whose composition progressively changes, and whose volume decreases through water reabsorption to less than 1%. Should it contain a substance with the ability to increase TER in a concentration-dependent manner, the 100-fold increase of electrical resistance from the proximal to the collecting duct would be accounted for.
GRANTS
This work was performed with research grants from the National Research Council of México (CONACYT). D. Flores-Benitez is recipient of student fellowship from the CONACYT.
ACKNOWLEDGMENTS
We express gratitude to Drs. A. Ponce and J. J. Gallardo for allowing unpublished experimental results with DLU from dog using whole cell patch-clamp techniques. We acknowledge the efficient and pleasant help of J. De Lorenz and E. del Oso.
【参考文献】
Acharya P, Beckel J, Ruiz WG, Wang E, Rojas R, Birder L, Apodaca G. Distribution of the tight junction proteins ZO-1, occludin, and claudin-4, -8, and -12 in bladder epithelium. Am J Physiol Renal Physiol 287: F305-F318, 2004.
Balda MS, Gonzalez-Mariscal L, Contreras RG, Macias-Silva M, Torres-Marquez ME, Garcia-Sainz JA, Cereijido M. Assembly and sealing of tight junctions: possible participation of G-proteins, phospholipase C, protein kinase C and calmodulin. J Membr Biol 122: 193-202, 1991.
Balda MS, Whitney JA, Flores C, González S, Cereijido M, Matter K. Functional dissociation of paracellular permeability and transepithelial electrical resistance and disruption of the apical-basolateral intramembrane diffusion barrier by expression of a mutant tight junction membrane protein. J Cell Biol 134: 1031-1049, 1996.
Beardmore JM, Richards RC. Concentrations of epidermal growth factor in mouse milk throughout lactation. J Endocrinol 96: 287-292, 1983.
Borok Z, Hami A, Danto SI, Lubman RL, Kim KJ, Crandall ED. Effects of EGF on alveolar epithelial junctional permeability and active sodium transport. Am J Physiol Lung Cell Mol Physiol 270: L559-L565, 1996.
Burrow CR, Wilson SJ, Wilson PD. Developmental regulation of ERB-B2 in human kidney organogenesis with persistent expression in autosomal dominant polycystic disease. Mol Biol Cell 8: 441a, 1997.
Carpenter G. Properties of the receptor for epidermal growth factor. Cell 37: 357-358, 1984.
Carpenter G. Employment of the epidermal growth factor receptor in growth factor-independent signaling pathways. J Cell Biol 146: 697-702, 1999.
Cereijido M. Electrical properties of Madin-Darby canine kidney cells. Fed Proc 43: 2230-2235, 1984.
Cereijido M, Anderson JM. Tight Junctions. Boca Raton, FL: CRC, 2001.
Cereijido M, Contreras RG, Shoshani L. Cell adhesion, polarity, and epithelia in the dawn of metazoans. Physiol Rev 84: 1229-1262, 2004.
Cereijido M, Herrera FC, Flanigan WJ, Curran PF. The influence of Na concentration on Na transport across frog skin. J Gen Physiol 47: 879-893, 1964.
Cereijido M, Shoshani L, Contreras RG. Molecular physiology and pathophysiology of tight junctions. I. Biogenesis of tight junctions and epithelial polarity. Am J Physiol Gastrointest Liver Physiol 279: G477-G482, 2000.
Cereijido M, Shoshani L, Contreras RG, Larre I. Na +,K + -ATPase, the star of epithelial net transports: what do we know about its own polarized distribution. Physiological Mini-Reviews 1: 1-10, 2005.
Chen MC, Goliger J, Bunnett N, Soll AH. Apical and basolateral EGF receptors regulate gastric mucosal paracellular permeability. Am J Physiol Gastrointest Liver Physiol 280: G264-G272, 2001.
Chen SP, Zhou B, Willis BC, Sandoval AJ, Liebler JM, Kim KJ, Ann DK, Crandall ED, Borok Z. Effects of transdifferentiation and EGF on claudin isoform expression in alveolar epithelial cells. J Appl Physiol 98: 322-328, 2005.
Claude P. Morphological factors influencing transepithelial permeability: a model for the resistance of the zonula occludens. J Membr Biol 39: 219-232, 1978.
Claude P, Goodenough DA. Fracture faces of zonulae occludentes from "tight" and "leaky" epithelia. J Cell Biol 58: 390-400, 1973.
Curran PF, Cereijido M. K fluxes in frog skin. J Gen Physiol 48: 1011-1033, 1965.
Dailey GE, Kraus JW, Orth DN. Homologous radioimmunoassay for human epidermal growth factor (urogastrone). J Clin Endocrinol Metab 46: 929-936, 1978.
Diamond JM. The mechanism of solute transport by the gall-bladder. J Physiol 161: 474-502, 1962.
Elson SD, Browne CA, Thorburn GD. Identification of epidermal growth factor-like activity in human male reproductive tissues and fluids. J Clin Endocrinol Metab 58: 589-594, 1984.
Fry DW, Kraker AJ, McMichael A, Ambroso LA, Nelson JM, Leopold WR, Connors RW, Bridges AJ. A specific inhibitor of the epidermal growth factor receptor tyrosine kinase. Science 265: 1093-1095, 1994.
Fukuyama R, Shimizu N. Detection of epidermal growth factor receptors and E-cadherins in the basolateral membrane of A431 cells by laser scanning fluorescence microscopy. Jpn J Cancer Res 82: 8-11, 1991.
Furuse M, Furuse K, Sasaki H, Tsukita S. Conversion of zonulae occludentes from tight to leaky strand type by introducing claudin-2 into Madin-Darby canine kidney I cells. J Cell Biol 153: 263-272, 2001.
Gallardo JM, Hernandez JM, Contreras RG, Flores-Maldonado C, Gonzalez-Mariscal L, Cereijido M. Tight junctions are sensitive to peptides eliminated in the urine. J Membr Biol 188: 33-42, 2002.
Gesualdo L, Di Paolo S, Calabro A, Milani S, Maiorano E, Ranieri E, Pannarale G, Schena FP. Expression of epidermal growth factor and its receptor in normal and diseased human kidney: an immunohistochemical and in situ hybridization study. Kidney Int 49: 656-665, 1996.
Gonzalez-Mariscal L, Contreras RG, Bolivar JJ, Ponce A, Chavez DR, Cereijido M. Role of calcium in tight junction formation between epithelial cells. Am J Physiol Cell Physiol 259: C978-C986, 1990.
Gonzalez-Mariscal L, Nava P, Hernandez S. Critical role of tight junctions in drug delivery across epithelial and endothelial cell layers. J Membr Biol 207: 55-68, 2005.
Gregory M, Dufresne J, Hermo L, Cyr D. Claudin-1 is not restricted to tight junctions in the rat epididymis. Endocrinology 142: 854-863, 2001.
Guerrero-Barrera AL, Garcia-Cuellar CM, Villalba JD, Segura-Nieto M, Gomez-Lojero C, Reyes ME, Hernandez JM, Garcia RM, de la GM. Actin-related proteins in Anabaena spp and Escherichia coli. Microbiology 142: 1133-1140, 1996.
Hirata Y, Orth DN. Epidermal growth factor (urogastrone) in human fluids: size heterogeneity. J Clin Endocrinol Metab 48: 673-679, 1979.
Holladay LA, Savage CR, Cohen S, Puett D. Conformation and unfolding thermodynamics of epidermal growth factor and derivatives. Biochemistry 15: 2624-2633, 2004.
Kim DC, Hanano M, Kanai Y, Ohnuma N, Sugiyama Y. Localization of binding sites for epidermal growth factor (EGF) in rat kidney: evidence for the existence of low affinity EGF binding sites on the brush border membrane. Pharm Res 9: 1394-1401, 1992.
Koefoed-Johnsen V, Ussing HH. The nature of the frog skin potential. Acta Physiol Scand 42: 298-308, 1958.
Kubota K, Furuse M, Sasaki H, Sonoda N, Fujita K, Nagafuchi A, Tsukita S. Ca 2+ -independent cell-adhesion activity of claudins, a family of integral membrane proteins localized at tight junctions. Curr Biol 9: 1035-1038, 1999.
Ladwein M, Pape UF, Schmidt DS, Schnolzer M, Fiedler S, Langbein L, Franke WW, Moldenhauer G, Zoller M. The cell-cell adhesion molecule EpCAM interacts directly with the tight junction protein claudin-7. Exp Cell Res 309: 345-357, 2005.
Lakshmanan J, Salido EC, Lam R, Fisher DA. Epidermal growth factor prohormone is secreted in human urine. Am J Physiol Endocrinol Metab 263: E142-E150, 1992.
Lipschutz JH, Li S, Arisco A, Balkovetz DF. Extracellular signal-regulated kinases 1/2 control claudin-2 expression in Madin-Darby canine kidney strain I and II cells. J Biol Chem 280: 3780-3788, 2005.
Masaki T, Stambe C, Hill PA, Dowling J, Atkins RC, Nikolic-Paterson DJ. Activation of the extracellular signal regulated protein kinase pathway in human glomerulopathies. J Am Soc Nephrol 15: 1835-1843, 2004.
Mattila AL. Human urinary epidermal growth factor: effects of age, sex and female endocrine status. Life Sci 39: 1879-1884, 1986.
Mattila AL, Perheentupa J, Pesonen K, Viinikka L. Epidermal growth factor in human urine from birth to puberty. J Clin Endocrinol Metab 61: 997-1000, 1985.
Mount CD, Lukas TJ, Orth DN. Characterization of a high-molecular-weight form of epidermal growth factor in an extract of human urine. Arch Biochem Biophys 255: 1-7, 1987.
Mroczkowski B, Reich M. Identification of biologically active epidermal growth factor precursor in human fluids and secretions. Endocrinology 132: 417-425, 1993.
Nava P, Lopez S, Arias CF, Islas S, Gonzalez-Mariscal L. The rotavirus surface protein VP8 modulates the gate and fence function of tight junctions in epithelial cells. J Cell Sci 117: 5509-5519, 2004.
Ohashi Y, Motokura M, Kinoshita Y, Mano T, Watanabe H, Kinoshita S, Manabe R, Oshiden K, Yanaihara C. Presence of epidermal growth factor in human tears. Invest Ophthalmol Vis Sci 30: 1879-1882, 1989.
Oka Y, Orth DN. Human plasma epidermal growth factor/beta-urogastrone is associated with blood platelets. J Clin Invest 72: 249-259, 1983.
Perheentupa J, Lakshmanan J, Fisher DA. Urine and kidney epidermal growth factor: ontogeny and sex difference in the mouse. Pediatr Res 19: 428-432, 1985.
Reyes JL, Lamas M, Martin D, del Carmen NM, Islas S, Luna J, Tauc M, Gonzalez-Mariscal L. The renal segmental distribution of claudins changes with development. Kidney Int 62: 476-487, 2002.
Rogers SA, Ryan G, Hammerman MR. Metanephric transforming growth factor-alpha is required for renal organogenesis in vitro. Am J Physiol Renal Fluid Electrolyte Physiol 262: F533-F539, 1992.
Saladik DT, Soler AP, Lewis SA, Mullin JM. Cell division does not increase transepithelial permeability of LLC-PK1 cell sheets. Exp Cell Res 220: 446-455, 1995.
Salido EC, Fisher DA, Barajas L. Immunoelectron microscopy of epidermal growth factor in mouse kidney. J Ultrastruct Mol Struct Res 96: 105-113, 1986.
Schaudies RP, Grimes J, Davis D, Rao RK, Koldovsky O. EGF content in the gastrointestinal tract of rats: effect of age and fasting/feeding. Am J Physiol Gastrointest Liver Physiol 256: G856-G861, 1989.
Shoshani L, Flores-Benítez D, Gonzalez-Mariscal L, Contreras RG. Regulation of tight junctions? functional integrity: role of a urinary factor, lipids and ouabain. In: Tight Junctions, edited by Gonzalez-Mariscal L. New York: Landes Bioscience and Springer Science+Buisness Media, LLC, 2006, p. 146-163.
Singh AB, Harris RC. Epidermal growth factor receptor activation differentially regulates claudin expression and enhances transepithelial resistance in Madin-Darby canine kidney cells. J Biol Chem 279: 3543-3552, 2004.
Skov OP, Nexo E, Poulsen SS, Hansen HF, Kirkegaard P. Renal origin of rat urinary epidermal growth factor. Regul Pept 10: 37-45, 1984.
Soler AP, Laughlin KV, Mullin JM. Effects of epidermal growth factor versus phorbol ester on kidney epithelial (LLC-PK1) tight junction permeability and cell division. Exp Cell Res 207: 398-406, 1993.
Sonoda N, Furuse M, Sasaki H, Yonemura S, Katahira J, Horiguchi Y, Tsukita S. Clostridium perfringens enterotoxin fragment removes specific claudins from tight junction strands: evidence for direct involvement of claudins in tight junction barrier. J Cell Biol 147: 195-204, 1999.
Tsukumo K, Nakamura H, Sakamoto S. Purification and characterization of high molecular weight human epidermal growth factor from human urine. Biochem Biophys Res Commun 145: 126-133, 1987.
Van Itallie C, Rahner C, Anderson JM. Regulated expression of claudin-4 decreases paracellular conductance through a selective decrease in sodium permeability. J Clin Invest 107: 1319-1327, 2001.
Van Itallie CM, Fanning AS, Anderson JM. Reversal of charge selectivity in cation or anion-selective epithelial lines by expression of different claudins. Am J Physiol Renal Physiol 285: F1078-F1084, 2003.
Vermeer PD, Einwalter LA, Moninger TO, Rokhlina T, Kern JA, Zabner J, Welsh MJ. Segregation of receptor and ligand regulates activation of epithelial growth factor receptor. Nature 422: 322-326, 2003.
Yu AS, Enck AH, Lencer WI, Schneeberger EE. Claudin-8 expression in Madin-Darby canine kidney cells augments the paracellular barrier to cation permeation. J Biol Chem 278: 17350-17359, 2003.
作者单位：1 Department of Physiology, Biophysics and Neurosciences, Center for Research and Advanced Studies (CINVESTAV), and 2 Laboratorio de Medicina, Facultades de Estudios Superiores-Iztacala y Cuautitlán, Universidad Nacional Autónoma de México, México City, México