INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

THE DIFFUSIONAL PERMEABILITY (P) to sucrose (P_suc) and to Na⁺ (P_Na⁺) of parietal pericardium has been recently measured in specimens taken from rabbits (36). The values of P_suc and P_Na⁺ (corrected for the effect of unstirred liquid layers) were 2.54 and 7.75 × 10⁵ cm/s, respectively. When phospholipids were added on the luminal side of the pericardium (where they are adsorbed; Refs. 12, 13), these values decreased to 0.71 and 3.93 × 10⁵ cm/s, respectively, in line with the decrease in P caused by adsorbed phospholipids in epithelia (11). Measurements of P_suc were also performed after the mesothelium was scraped away from the specimens: these data provided P_suc of the connective tissue. Because the mesothelium and the connective tissue are placed in series, their resistances to diffusion add up. Therefore, 1/P_suc of the whole specimen minus 1/P_suc of the scraped sample yields 1/P_suc of the mesothelium alone. P_suc of the mesothelium was 2.92 and 0.74 × 10⁵ cm/s without and with phospholipids, respectively. Hence, most of the resistance to diffusion of the pericardium is provided by the mesothelium, although the connective tissue is 35 times thicker (36). In the experiments with phospholipids, P_suc of the mesothelium seems similar to P_suc of the leaky epithelium of the renal proximal tubule, 0.69 × 10⁵ cm/s (14).

In the present research we measured P of the parietal pericardium of rabbits, without and with phospholipids, to other molecules (mannitol, Cl, and water). P to the above molecules and P_Na⁺ were then measured in specimens in which the mesothelium had been scraped away to get data for the connective tissue, and, hence, to compute those for the mesothelium alone (see above). From the previous and the present data (except that of water, because of its diffusion through the cellular membrane; Refs. 4, 23, 30, 34) we tried to determine the equivalent radius of the "pores" of the intercellular junctions of the mesothelium with two approaches. The first is a graphical analysis based on the comparison between the experimental values of P and the theoretical relationships between P and the molecular radius for free diffusion and restricted diffusion through paracellular pores. The second is that used by Preisig and Berry (23).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The experiments were performed in 96 giant rabbits (body weight 5-7 kg, age 7-11 mo). The animals were anesthetized with a solution (2 ml/kg iv) containing pentobarbital sodium (Sigma Chemical, 10 mg/ml) and urethan (Sigma Chemical, 250 mg/ml) and placed supine on a tilting board, 20°, head up. The trachea was cannulated to ensure adequate ventilation during the preliminary surgical procedure, and airflow and tidal volume were recorded on a 7418 Hewlett-Packard thermopaper oscillograph.

Specimen collection and preparation. Collection and preparation of the specimens of the sternal part of the parietal pericardium (which is essentially free of stomas; Ref. 27) were performed with the procedure previously described (36), which minimizes manipulation and air exposure of the mesothelium. Briefly, after the rabbit was killed by an overdose of anesthetic, a segment of sternum was removed, leaving undamaged the underlying parietal pericardium. After the larger fat patches projecting from the pericardium were removed, a roughly rectangular specimen of the latter (~3 × 2 cm) was hooked and excised, while an albumin-Ringer solution was being poured on the pericardium to prevent air exposure of the mesothelium. The specimen was never stretched during removal, and the whole procedure was completed within 4 min of the death of the animal. The specimen, covered by the albumin-Ringer solution, was pinned with its interstitial side facing upward, at its in situ length and width, to a layer of Sylgard (Dow Corning) adherent to the bottom of a petri dish. The solution was bubbled continuously with a 95% O₂-5% CO₂ gas mixture (22). Small vessels, fat patches and, when present, blood clots were removed from the interstitial side of the specimen until a transparent area of ~1 × 1.5 cm was obtained; the mesothelium of the central part of the specimen was never touched. The cleaning procedure took 20-25 min. In 34 experiments (to assess connective tissue permeability, see below), after this procedure the specimen was turned and pinned with its luminal side facing upward, the albumin-Ringer solution was removed, and the mesothelium was gently scraped away with the blade of a scalpel (35, 36).

Solutions. The composition of the Ringer solution used during specimen collection and preparation, as well as during the experiments was (in mM) 139 Na⁺, 5 K⁺, 1.25 Ca²⁺, 0.75 Mg²⁺, 119 Cl, 29 HCO₃, and 5.6 D-glucose. Rabbit albumin (Sigma Chemical, 0.5 g%) was added to maintain normal permeability (7). The following phospholipids were used (36): 50% dipalmitoyl phosphatidylcholine (367 µg/ml), 32% dipalmitoyl phosphatidylethanolamine (235 µg/ml), and 18% sphingomyelin (132 µg/ml). Radioactive markers were added at the following specific activities: 0.5 µCi/ml for ³H₂O (Sigma Chemical), ²²Na⁺ (Amersham), and [³H]mannitol (ICN); and 0.2 µCi/ml for ³⁶Cl (NEN Life Science Products). Overall concentrations of the solutes (i.e., labeled plus unlabeled) in the donor chamber were 0.119 mmol/ml for Cl; 0.139 mmol/ml for Na⁺; and 1.9 × 10⁸ mmol/ml for mannitol. In all experiments with [³H]mannitol, the same concentration of unlabeled mannitol was added to the recipient chamber.

Experimental protocols. A first incubation period of 30 min was allowed for tissue recovery, temperature equilibration, and initial phospholipid adsorption when scheduled. At the end of this incubation period, both chambers were simultaneously emptied and the recovered liquid was stored for determination of background radioactivity (see below). Simultaneous refilling of the chambers was immediately made with 4 ml of labeled solution in the donor and 4 ml of unlabeled solution in the recipient chamber. A second incubation period was allowed for attaining equilibrium of tracers between the solution in the donor chamber and the specimen, and to continue phospholipid adsorption when scheduled. The duration of this period ranged from 30 min (mannitol) to 15 min (water) according to the time required to reach steady-state flux. At the end of this period, a sample of 50 µl was withdrawn from the donor chamber, whereas all the liquid was removed from the recipient chamber, which was immediately refilled with a volume of fresh unlabeled solution equal to that present in the donor chamber after removal of the 50-µl sample (3.95 ml). The time required for liquid withdrawal and replacement was 3-4 s. At the end of this procedure, the first measurement period started. The duration of the measurement period was different according to the molecule tested and the specimen used (intact or scraped), to ensure that isotope concentration in the recipient chamber remained negligible (<2%) relative to that in the donor chamber. Indeed, this is the requisite to prevent isotope backdiffusion, thus allowing measurement of unidirectional (rather than net) fluxes. Measurement period duration was 20 min in all experiments with mannitol and 5 min in all experiments with ³H₂O; with ³⁶Cl, it was 15 min with intact and 5 min with scraped specimens; with ²²Na⁺ (only scraped specimens), it was 8 min. The procedure described at the end of the second incubation period was repeated at the end of the first measurement period to perform a second measurement period.

Measurement of P. The samples of liquid withdrawn from each chamber at the end of the first and second measurement periods were treated as previously described (36), and -activity was determined as counts per minute (cpm) in a liquid scintillation spectrometer (Minaxi  Tri-Carb 4000, Packard Instruments). After correction for background radioactivity, average values were expressed as counts per minute per milliliter to provide values proportional to isotope concentration in a given chamber. Checks for constant isotope concentration in the donor chamber, and for negligible isotope concentration in the recipient relative to the donor chamber at the end of each measurement period, were performed as previously described (36). Because isotope concentration in the recipient chamber was nil at the beginning of each period, the unidirectional flux () of a given molecule is given by  = (^*C'R CD VR)/(^*CD A t), where ^*C'R is the isotope concentration in the recipient chamber at the end of a measurement period; CD is the overall concentration of the solute (i.e., labeled plus unlabeled) in the donor chamber (see below); VR is the volume of the solution in the recipient chamber; ^*CD is the isotope concentration in the donor chamber; A is the surface area of the window; and t is the duration of each measurement period. P to a given molecule was then obtained, according to Fick's law, from P = /CD. The values of P thus obtained were corrected for the effect of liquid unstirred layers (USL) close to the membrane, by using the formula of resistances in series (3): 1/P_cor = (1/P_meas)  (d_liq/D), where P_cor is the corrected P, P_meas is measured P, d_liq is the overall USL thickness, and D is the diffusion coefficient of the solute in water at 37°C. The value of P_cor has been used for further calculations throughout this study. The thickness of the USL facing the luminal side was assumed to be 70 µm and that of the USL facing the interstitial side to be 100 µm (3): the assumed value of d_liq was therefore 170 µm in the experiments on intact specimens and 200 µm in those on scraped specimens; the D values used are reported in Table 1. The P values obtained in experiments on integer specimens provided the P of the pericardium (P_per), whereas those obtained in experiments on scraped specimens provided that of the pericardial connective tissue (P_con). Because the mesothelium and the connective tissue are placed in series, their resistances to diffusion (1/P) add up: therefore, P of the mesothelium (P_mes) was computed by using the formula of series resistances: (1/P_mes) = (1/P_per)  (1/P_con).

Estimation of equivalent pore radius of intercellular junctions. The equivalent pore radius of the intercellular junctions of the mesothelium was determined in two ways. The first is a graphical analysis based on the comparison between the experimental values of P and the theoretical relationships between P and molecular radius (a) for free diffusion and restricted diffusion through paracellular pores. Under conditions of free diffusion, i.e., when the pore radius (r) is at least 100 times > a

(1)

where A is the area of the membrane, and A_p/l is the overall cross-sectional area per unit pathlength of the pores (20). Under conditions of free diffusion in water, the relationship between D and a is provided by the Stokes-Einstein equation: D = (RT)/(N6a), where R is the gas constant, T is the absolute temperature,  is the viscosity of water, and N is the Avogadro number. Hence, by substituting D in Eq. 1, one obtains the theoretical P-a relationship for free diffusion: P = (RT A_p)/(A l N6a).

(2)

(3)

(4)

Measurement of thickness. The thickness of the specimens was determined at the end of the experiments by focusing on two tantalum dust particles (particle size  5 µm, Sigma Chemical) situated approximately along the same axial line on either side of the specimen, as previously described (36). The measurement was repeated in 10 different sites of the specimen. The mean of these measurements was taken as the average thickness of the specimen.

Statistics. Data are expressed as means ± SE. Statistical significance of differences between groups was assessed by analysis of variance.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The mean P_w, P_Cl, and P_man values of the sternal part of the parietal pericardium, measured in the experiments without and with phospholipids in the solution facing the luminal side of the specimen, are reported in Table 2, along with P_Na⁺ and P_suc obtained in the previous research (36). The same values corrected for the effect of USL (see METHODS) are also reported in Table 2. In the experiments without phospholipids, the interstitium-lumen Cl flux (51.0 ± 7.7 µmol · h¹ · cm²) was not significantly lower than that in the opposite direction (55.2 ± 7.4 µmol · h¹ · cm²). Because we previously found that also Na⁺ flux was similar in both directions (Table 1 in Ref. 36), these data suggest that, if an active transport occurs, it is negligible relative to diffusion, at least under our experimental conditions. In the experiments with phospholipids, P_man decreased markedly, in line with the previous findings on P_suc and P_Na⁺ (36); instead, the decrease in P_Cl is not significant. We do not know the cause of this finding (see DISCUSSION).

The mean values of P_w, P_Cl, P_Na⁺, and P_man measured in the experiments in which the mesothelium was scraped away from the specimen are reported in Table 3, along with that of P_suc obtained in the previous research (36). The same values corrected for the effect of USL (see METHODS) are also reported in Table 3. These values provide P of the connective tissue of the parietal pericardium. The thickness of the scraped specimens was 67.3 ± 1.3 µm, whereas that of the unscraped ones was 73.4 ± 1.6 µm (being 74.2 ± 2.3 µm in the experiments without phospholipids and 72.1 ± 2.1 µm in those with phospholipids). Because the mesothelium is ~2 µm thick (27, 32), the connective tissue of the specimen is ~35 times thicker than the mesothelium, in line with our previous finding (36). Moreover, because the thickness of the scraped specimens was 6.1 µm smaller (P < 0.01) than that of the unscraped ones, only a few micrometers of connective tissue were removed by scraping the specimen.

The mean values of P to the various molecules of the mesothelium alone (see METHODS), without or with phospholipids, are reported in Table 3. The value of P_w in the experiments without phospholipids must be taken cautiously because small changes in P_w of the pericardium and/or of the connective tissue lead to marked changes in P_w of the mesothelium. The resistance to diffusion of Cl, Na⁺, mannitol, and sucrose through the mesothelium without phospholipids is 3.3, 4.4, 6.0, and 7.3 times greater, respectively, than that through the connective tissue, although the latter is ~35 times thicker than the former. With phospholipids the above values become 4.1, 9.6, 13.4, and 29.0, respectively. On the other hand, the resistance to diffusion of water through the mesothelium without and with phospholipids is smaller than that through the connective tissue, being 34 and 78%, respectively. The different behavior of water is probably due to its marked diffusion through the cellular membrane, as in epithelia (4, 23, 30, 34), whereas the above hydrophilic solutes (see METHODS) diffuse only through the intercellular junctions.

The mean values of P of the mesothelium to the various molecules used are plotted as a function of the molecular radius (a) in Fig. 1. In the same diagram are drawn the theoretical P-a lines for free diffusion and restricted diffusion through the paracellular pores of a given radius (r). The free-diffusion line is computed with an A_p/l value of 3.04 cm, which is the mean between those obtained from P_Cl (3.44 cm) and P_Na⁺ (2.64 cm) in the experiments without phospholipids (see METHODS). The lines for restricted diffusion are computed from the free-diffusion values times the Renkin function for a given pore radius (r; see METHODS). In the experiments without phospholipids, P_man and P_suc fit a line corresponding to a pore radius of 6 nm; in the experiments with phospholipids, P_man and P_suc fit lines corresponding to 1.9 and 1.6 nm, respectively. In both kinds of experiments, P_w is much higher than the corresponding point on the free-diffusion line, suggesting a marked diffusion of water through the cellular membrane, like in epithelia (4, 23, 30, 34).

View larger version (17K): [in this window] [in a new window]   Fig. 1.   Diffusional permeability (P) of mesothelium of parietal pericardium to various molecules plotted vs. molecular radius (a) in experiments without phospholipids (open symbols) and with phospholipids in solution facing luminal side (solid symbols). Theoretical P-a relationships for free diffusion (FD) and for restricted diffusion with a given "pore" radius (r) are also shown (see text). Logarithmic scale on ordinate has been used for graphical reasons and because it has been formerly used in similar diagrams (16, 22).

The value of the ratio (P_man/D_man)/(P_suc/D_suc) for the mesothelium was 1.05 and 1.83 in the experiments without and with phospholipids, respectively. These values, which enable the computation of the paracellular pore radius without using A_p/l (see METHODS; Ref. 23), correspond to r values of 7.8 and 1.1 nm, respectively. Therefore, the values of r obtained with the two approaches agree.

The mean values of P of the connective tissue to the various molecules used are plotted as a function of molecular radius in Fig. 2. In the connective tissue of the pericardium, which is relatively loose, a condition of free diffusion should apply to all molecules used, because, in a loose connective tissue like the subcutaneous one, the hydraulic radius of the pores (which is smaller than the actual radius of the pores) has been found to be ~20 nm (18). Therefore, assuming that the analysis of diffusion through porous membranes (20) may be applied in a first approximation to connective tissue matrix, in Fig. 2 is also drawn the theoretical P-a relationship for free diffusion computed with an A_p/l value of 12.4 cm, which is the mean A_p/l value of all molecules used. It should be considered, however, that whereas the values of A_p/l for water, Cl, and Na⁺ were 11.4, 11.4, and 11.6 cm, respectively, those for mannitol and sucrose were 12.9 and 15.1 cm, respectively. The reason for the higher values for mannitol and sucrose is not clear.

View larger version (12K): [in this window] [in a new window]   Fig. 2.   P of connective tissue of parietal pericardium to various molecules plotted vs. molecular radius (a) and theoretical P-a relationships for free diffusion (see text).

At variance with the mesothelium, in the connective tissue P_w is close to the P-a line (Fig. 2) because the connective tissue is nearly cell free and, therefore, the contribution of water diffusion through the cell membrane is negligible. Moreover, P_w in the connective tissue is lower than in the mesothelium, at variance with P to the solutes. This seems due to water diffusion through the mesothelial cells, which provides a high P_w in the mesothelium. Indeed, without water diffusion through the cellular membrane P_w of the mesothelium (see left end of free-diffusion line in Fig. 1) would be about one-third that of the connective tissue. P_Cl and P_Na⁺ are close to the P-a line because the negative electric charges of the macromolecules of the connective matrix do not seem to affect the diffusivity of small ions (19). Because D_man and D of glucose are the same (5), the value of P_man through the connective tissue, as found in the present research, provides the diffusion rate of glucose through the connective tissue: ~2 µm/s.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

It has been recently shown that P_suc of specimens of parietal pericardium of rabbits obtained in such a way as to minimize mesothelial damage (36) is ~10 times smaller than that of stripped specimens of visceral pleura of sheep (16) and of visceral and parietal pleura of dogs (22), despite the similarity of intercellular junctions in the mesothelium of pericardium and pleura of various species (1, 8, 15, 32). Most of this difference has been ascribed to the changes undergone by the pleural mesothelium during the process of getting the specimens, because morphological researches have shown that mesothelial cells lose mutual contact when they are irritated (6), they are easily detached during handling of the tissue (8), and their intercellular junctions widen on simple exposure to air (26, 28). Moreover, no albumin was added to the Ringer solution used in the experiments by Kim et al. (16) and Payne et al. (22), and it has been shown that lack of albumin increases the permeability of the capillary endothelium (7). Two of these explanations have been supported by the finding that P_suc of parietal pericardium increased four times when air exposure was not prevented and 0.5% albumin was not added to the Ringer solution (36).

P_w of the parietal pericardium measured in the present research (41 × 10⁵ cm/s; Table 2) is 20% smaller than that of stripped visceral pleura of sheep (51 × 10⁵ cm/s; Ref. 16) and ~67% smaller than that of stripped visceral and parietal pleura of dogs (122 and 127 × 10⁵ cm/s, respectively; Ref. 22). Therefore, the difference in P_w is small relative to that in P_suc (see above). The causes of the smaller difference in P_w appear to be the following. 1) The widening of intercellular junctions produces a relatively smaller increase in P of a molecule that diffuses also through cellular membrane. 2) The experimental procedure followed by Kim et al. (16) and Payne et al. (22) to measure P does not prevent backdiffusion of the labeled molecule in the Ussing chamber: because backdiffusion of labeled water may be substantial, P_w is underestimated in both researches.

In line with the high value of P found in their stripped specimens of visceral pleura (which lacks stomas), Kim et al. (16) and Payne et al. (22) computed an equivalent pore radius of 80 and 75 nm, respectively. In the present research the equivalent radius of the paracellular pores of the mesothelium of the parietal pericardium computed with two approaches was 6-7.8 nm in the experiments without phospholipids and 1.7-1.1 nm in those with phospholipids. Considering that our specimens also underwent some handling, and, therefore, may have been damaged (though markedly less so than the stripped specimens of pleura), it seems likely that under physiological conditions the equivalent radius of the paracellular pores of the mesothelium is ~5 nm or even smaller. This value agrees with those found in leaky epithelia: 4 nm in rabbit gallbladder (29), 1.4 nm in rat proximal tubule (23), and 5 nm in rat ileum (21). It is also similar to the equivalent pore radius of the endothelium of muscle capillaries (~5 nm; Ref. 17). An equivalent pore radius of 6 nm has been estimated in cat peritoneum by Rippe et al. (25) from measurements of osmotic water conductance. The present research has been addressed to the determination of the equivalent pore radius of the intercellular junctions of the mesothelium. It might be that a few large pores also occur in the intercellular junctions of the mesothelium, as in the endothelium of most capillaries (17). Moreover, a few stomas could also be present, although the region of parietal pericardium investigated should be essentially free of them (27). Finally, although the morphological features of the mesothelium are similar in the pleura, pericardium, and peritoneum (1, 8, 15, 26, 32), it could be that small functional differences occur among these serous membranes.

Under conditions of free diffusion, the relative pore area (A_p/A) is given by Pl/D (see METHODS, Eq. 1). For the paracellular pores of the mesothelium, A_p/A may be roughly computed by assuming a nearly free diffusion of Cl and Na⁺ through these pores (see above) and a pathlength of the pores similar to the thickness of the mesothelium in the region of intercellular junctions (~1.5 µm; Refs. 27, 32), as has been done for capillary endothelium (5, 20). Therefore, by using the mean between P_Cl and P_Na⁺ of the mesothelium in the experiments without phospholipids (Table 3) and the mean between D_Cl and D_Na⁺ (Table 1), A_p/A = (13.7 × 10⁵ cm/s) × (1.5 × 10⁴ cm)/(2.2 × 10⁵ cm²/s) = 9.3 × 10⁴, that is, ~0.1%. However, this is an overestimation because the sieving pores correspond to the tight junctions, and their length, considering the tortuosity of the path, should be ~0.3 µm (32). Hence, A_p/A = (13.7 × 10⁵ cm/s) × (0.3 × 10⁴ cm)/(2.2 × 10⁵ cm²/s) = 1.9 × 10⁴, that is, ~0.02%. The relative pore area of the connective tissue may be computed by taking the mean P/D value of the molecules used (24.9 cm¹) and assuming that, because of tortuosity, the path length of the pores is 43% greater than tissue thickness (18), which was 67.3 µm in our scraped specimens (see RESULTS). Hence A_p/A = (24.9 cm¹) × (96.1 × 10⁴ cm) = 0.24, that is, ~24%. Therefore, the relative pore area in the connective tissue is at least two orders of magnitude greater than that in the paracellular path of the mesothelium (see above).

The finding that P_Cl of the pericardium (Table 2) or of the mesothelium (Table 3), at variance with P to the other solutes used, is not significantly decreased by adding phospholipids to the solution facing the luminal side remains so far unexplained. This finding suggests that adsorption of phospholipids to the luminal side of the mesothelium exerts little hindrance to Cl diffusion, although they are not positively charged in the physiological range of pH (2). This intriguing finding is being investigated in a study of the electrical resistance of the mesothelium.