Editors' Highlight pubs.acs.org/ac
Deuterium Oxide Dilution: A Novel Method to Study Apical Water Layers and Transepithelial Water Transport Daniel Neubauer,† Jonas Korbmacher,‡ Manfred Frick,‡ Johanna Kiss,‡ Melanie Timmler,‡ Paul Dietl,‡ Oliver H. Wittekindt,*,‡ and Boris Mizaikoff*,† †
Institute of Analytical and Bioanalytical Chemistry, University of Ulm, Albert-Einstein-Allee 11, 89081 Ulm, Germany Institute of General Physiology, University of Ulm, Albert-Einstein-Allee 11, 89081 Ulm, Germany
‡
S Supporting Information *
ABSTRACT: Lung epithelia regulate the water flux between gas filled airways and the interstitial compartment in order to maintain organ function. Current methodology to assess transepithelial water transport is limited. We present a D2O dilution method to quantify submicroliter volumes of aqueous solutions on epithelial cell layers. Evaluating D2O/H2O mixtures using mid-infrared (2−25 μm) attenuated total reflection (ATR) spectroscopy, with a resolution of 0.06% vol/ vol change, corresponding to 24 nL, was achieved. Using this method, we demonstrate that water transport across NCI-H441 lung epithelial cell layers and apical surface liquid (ASL) volumes are coupled to dexamethasone dependent amiloride-sensitive ion transport. However, contrary to current dogma, electrogenic transport is not rate-limiting for water transport. This clearly indicates the need to directly assess net water rather than ion transport across epithelial cell layers. The presented D2O dilution method enables such direct and quick quantification of transepithelial water transport with high resolution.
T
resulting mixtures was analyzed using attenuated total reflection (ATR) techniques.23,24 This D2O dilution method enables one to precisely analyze ASL volumes and water transport across NCI-H441 lung epithelial cell layers requiring a minute sample volume.
he apical surface of lung epithelia is lined by a thin water layer called the apical surface liquid (ASL). The thickness of this layer is adjusted by epithelial transport.1 Transepithelial water transport is tightly coupled to electrogenic ion transport2 and, therefore, is frequently indirectly estimated on the basis of electrophysiological measurements. Even though ion transport drives water flux, it is well-known that both transport processes may be separately regulated.2−4 Thus, direct water transport measurements are essential to understand epithelial function. To date, such measurements predominantly use marker substances5,6 or gravimetry2 to investigate water flux across cell barriers. Gravimetric measurements are limited by unknown amounts of adhering liquid. High molecular weight markers, for which epithelial cell layers are almost impermeable,7 may not diffuse into volumes near the plasma membrane8 or the pericilliar space on airways.9 Fluorescent molecules can change their fluorescence spectrum or optical yield by changes of the pH10,11 or by the presence of oxidants12 and may be quenched by cellular metabolites.13 Such interferences limit the number of applicable fluorescent dyes and affect volume and water transport measurements, thus only allowing rough estimations rather than an absolute quantification of aqueous volumes or volume changes. As previously reported, H2O/D2O mixtures may readily be analyzed by mid-infrared (MIR) spectroscopy;14−18 therefore, D2O was successfully used to trace water within biological systems.19−22 Herein, for accurate quantification, the aqueous volumes on top of epithelial cell layers were diluted by defined volumes of D2O solutions, and the H2O content of the © 2013 American Chemical Society
■
EXPERIMENTAL SECTION Sample Preparation for Apical Surface Liquid (ASL) Volume Studies. ASL volumes were expected not to exceed 10 μL on top of the used cell layers herein. After basolateral solution was removed, ASL was diluted by 40 μL of isotonic NaCl solution in D2O. To get access to the cytoplasm, 0.05% Triton X-100 (Sigma, Germany) was added to the isotonic D2O solution. Sample Preparation for Water Transport Measurements. Basolateral medium was exchanged by 400 μL of medium equilibrated with 5% CO2 at 37 °C. Twenty-five microliters of isotonic NaCl solution was added to the apical side. Maltose (Sigma, Germany) was added to either medium or apical solutions in order to adjust osmotic gradients. Osmolality of solutions and media was measured using an OM801 Osmometer (Vogel, Gießen, Germany). To avoid apical evaporation, nonused wells as well as space between the wells were filled with isotonic NaCl solution. Evaporation was estimated using 3 to 4 silicone sealed filter inserts placed Received: January 26, 2013 Accepted: April 5, 2013 Published: April 5, 2013 4247
dx.doi.org/10.1021/ac4002723 | Anal. Chem. 2013, 85, 4247−4250
Analytical Chemistry
Editors' Highlight
Several studies26 have attempted to measure transepithelial water transport. However, to the best of our knowledge, no method qualifies as a reference to establish an accurate approach for analyzing transepithelial water transport. For verifying the utility of this method, a model epithelium was established on the basis of NCI-H441 (H441) cells cultivated at different concentrations of dexamethasone [Dex], which is known to induce the expression of epithelial sodium channels (ENaC). High expression leads to an increased electrogenic ion transport followed by an osmotic diffusion of water. These cell layers facilitate the correlation of ion transport capacity with the water transport rate (Figure S-2, Supporting Information). Images obtained via confocal microscopy illustrate that the ASL thickness significantly varies for different surface regions (Figure 2a,b), thus reflecting a nonuniform volume distribution.
equally distributed on each plate. Compounds were added at given concentrations to both compartments. If not mentioned otherwise, remaining apical volume was measured by adding 25 μL of isotonic NaCl solution in D2O after 16 h to the apical compartment. Apical solutions were mixed by pipetting. The entire D2O/H2O mixture was transferred to a reaction tubule. To avoid any exchange with atmospheric humidity, tubules were tightly sealed until measurements were performed. For calibration, a concentration series with 15%, 25%, 40%, 50%, and 65% (vol/vol) of H2O in D2O was prepared. See Supporting Information for ASL measurements. Infrared Spectroscopic Analysis. All measurements were performed using a Vertex 70 FT-IR spectrometer equipped with a BioATRCell-II unit and a liquid nitrogen cooled MCT detector (Bruker Optics, Ettlingen, Germany). Data acquisition and processing was performed using an OPUS 6.5 software package (Bruker Optics, Ettlingen, Germany). The BioATRCell-II sample chamber was preconditioned with a 10 μL sample aliquot. Then, 15 μL of sample was added, and the measurement (100 scans, 4000 to 400 cm−1, spectral resolution: 4 cm−1) was started immediately to reduce influence from proton−deuterium exchange with air moisture. Prior to each set of samples, a background spectrum of the empty ATRunit was collected and a calibration set was measured.
■
RESULTS AND DISCUSSION The analysis of IR ATR spectra of H2O in D2O (Figure S-1, Supporting Information) provided for a linear calibration function evaluating the H−O stretching (∼3400 cm−1), D−O stretching (∼2500 cm−1), and D−O−D bending band (1210 cm−1)25 area for H2O concentrations ranging from 1% to 25%, and from 15% to 65% (vol/vol) (Figure 1). The maximum
Figure 2. (a, b) Apical surface liquid (ASL) visualized in z-stacks, as acquired by confocal microscopy. The ASL thickness varied for randomly selected areas at a representative H441 cell layer. Scale bar = 30 μm. Filter layer is localized between dotted lines. (c) Volume measured at control conditions (isotonic D2O) and the presence of 0.05% Triton X-100 (Triton). (d) ASL volumes analyzed using the D2O dilution method. Cells were cultivated at given [Dex] (mol/L) in air/liquid configuration.
The D2O dilution method provides total ASL volumes independent of such thickness inhomogeneities. Using the D2O dilution method, the ASL is diluted with a defined volume of 0.9% (w/vol) NaCl solution in D2O (Figure S-3, Supporting Information), and the apical cell surface is exposed to D2O for less than 20 s. Previous studies have demonstrated a rapid exchange between the aqueous cytoplasm and extracellular D2O,19,22 and it is anticipated to determine extra- and intracellular volumes but not exclusively ASL. To test this hypothesis, isotonic D2O solution containing the detergent Triton X-100 was added to the apical surface of H441 monolayers in order to permeabilize the cells. The measured volume then included the cytoplasm, and was found to (Students t test, p = 0.018) increase by 0.4 μL compared to the volume measured by D2O solution lacking any detergent (Figure 2c). This volume difference would correspond to a cell layer thickness of 12 μm, which agrees with confocal microscopy studies (Figure 2a,b). While these experiments do not exclude D2O exchange with the cytoplasm, it is clearly evident that the ASL accounts for the major volume analyzed during these experiments. This method (Figure 2d) revealed that ASL volumes remained stable at 1.2 μL (SD = 0.08 to 0.2 N = 6), if the cell layers were cultivated at [Dex] of 3 nM and above.
Figure 1. Typical H2O calibration series (a) 15% to 65% and (b) 1% to 25% (vol/vol). Band area integration values in arbitrary units for H−O stretching, D−O stretching, and D−O−D bending vibrations. See Supporting Information for details.
volume resolution was calculated to be 0.06% (vol/vol) (4σ, R = 1). This corresponds to a volume of 24 nL (40 μL of D2O dilution volume) and a water layer thickness of 0.7 μm (epithelia surface area was 33 mm2). In contrast to methods analyzing traces of marker substances, the D2O dilution method herein is directly based on main components and utilizes their absorption features for molecular identification. Thus, this method is insensitive toward low concentrations of dissolved contaminants originating from, e.g., the cell culture. Both the high volume resolution and the inherent robustness of the analytical technique render this method ideal for quantifying aqueous volumes within biological systems. However, for measurements involving higher contents of other constituents than water and deuterium oxide, it is recommended to prepare separate calibration series. 4248
dx.doi.org/10.1021/ac4002723 | Anal. Chem. 2013, 85, 4247−4250
Analytical Chemistry
Editors' Highlight
However, they increased to a maximum of 6.2 μL (SD = 1,3 μL, N = 6 for [Dex] = 0 nM), if the cell layers were cultivated at [Dex] below 3 nM. Dex was previously shown to affect tight junction formation and epithelial barrier function.27−29 The H441 epithelia studied herein were cultivated at stable air/ liquid interfaces, and therefore, the increased ASL volumes at low [Dex] are not caused by leakage rather than by regulatory mechanisms, which prevent the apical surface from desiccation.30,31 The measured ASL volume of 1.2 μL corresponds to 36 μm of ASL height, which is in line with previous studies on lung epithelial cell layers.32 To measure water transport, 25 μL of isotonic NaCl solution was added to the apical side of H441 epithelia (Figure S-4a, Supporting Information). After incubation for 16 h, the remaining volume was measured by the D2O dilution method. Dexamethasone (Dex) induces an increase in amiloridesensitive ion transport, as well as in transepithelial water transport rate (Figure 3a,b). The increase of apical volume during incubation observed at cell layers cultivated at [Dex] of 3 nM and below may indicate water secretion. However, it has to be considered that ASL volumes themselves add some uncertainty to these measurements. The measurement of ASL volumes by the D2O dilution method causes the presence of D2O at the apical surface, which disables its determination prior to water transport. Especially in cell layers cultivated at low [Dex], the large ASL volumes would result in an overestimation especially of secretory water transport. Although ion transport (measured as ISC) drives water flux, the observed water transport rate does not exceed a certain limit regardless of further increased ISC (Figure 3c). Evidently, ISC is not rate limiting for the water transport. Decoupling of the water transport from ion transport is assumed to be a regulatory mechanism of transepithelial transport.3,4 A similar mechanism is most likely involved in regulating water transport across H441 epithelia. Consequently, these observations corroborate the need of direct water transport measurements for evaluating epithelial transport function. Amiloride blocked ISC in H441 cell layers almost completely, whereas water transport was only blocked by 70% with a 6-fold decreased potency (Figure 3d,e). The fact that water transport is independent from amiloride to a certain extent has already been explained in previous studies.33−36 Low amiloride concentrations would affect ISC but would not affect the water transport rate. Only if amiloride concentrations exceed values at which ISC becomes rate limiting for the water transport both mechanisms would be blocked. Water is exchanged between the apical and the basolateral compartment via diffusion. A vectorial transport, which results in a volume change of fluid in either compartment, depends on an active transport of osmolytes. This may be simulated by the application of an osmotic gradient between the basolateral and the apical compartments (Figure 3f). As expected, water secretion was observed for a basolateral-to-apical osmotic gradient of 40 mOsm. An inverted gradient accordingly resulted in water resorption. As demonstrated, amiloride inhibits almost 70% of water resorption. According to its ability to inhibit Na+ resorption, amiloride increases the secretory water flux in the presence of a basolateral-to-apical osmotic gradient and decreases resorptive water flux in the presence of an inverted osmotic gradient. In both cases, the shifted water volume in the presence of amiloride was sufficient to balance osmolarity between basolateral and apical solutions. The calculated water transport rates given in Figure 3f do not account for the time dependent decay of the osmotic gradient.
Figure 3. (a) Dependence of the amiloride-sensitive electrogenic transport measured as short circuit current (ISC), across NCI-H441 cell layers for selected dexamethasone concentrations in mol/L applied during cultivation. (b) Dependence of water transport/flux on dexamethasone concentrations. For measurements, the apical side of the cell layers was supplied with 25 μL of isotonic NaCl solution and incubated for 16 h. Residual water was then quantified by the D2O dilution method, and the flux was calculated. Positive values represent resorptive water transport. (c) Water transport rate plotted against amiloride sensitive ISC. (d, e) Response of ISC and water transport to amiloride, if cultivated in the presence of 300 nM dexamethasone. (d) Relative ISC was calculated as normalized ISC to ISC measured at control conditions and plotted against amiloride concentrations in mol/L. (e) Water transport rates plotted against amiloride concentrations in μmol/L. See Supporting Information for details. (f) Effect of osmotic gradient on water flux. +40 mOsm = increased apical osmolality by 40 mOsm manitol and −40 mOsm = increased basolateral osmolality by 40 mOsm manitol. Amil = application of 100 μM amiloride to the apical and basolateral compartment. Shifted volumes were measured after a 5 h incubation time, and transport rate was calculated accordingly.
■
CONCLUSION A simplified cellular model was used to establish the D2O dilution method for studying transepithelial water transport and for demonstrating the superiority of this method compared to currently applied marker-based methods. It should be noted that water exchange between compartments separated by epithelia may also occur via diffusion.21 Prior to apical D2O addition, the basolateral solution was removed from the cell layers. However, D2O diffusion may occur due to adhering liquid. Therefore, the D2O exposure time was kept
