Pleural mesothelium lubrication after hyaluronidase, neuraminidase or pronase treatment

Respiratory Physiology & Neurobiology 188 (2013) 60–65

Contents lists available at SciVerse ScienceDirect

Respiratory Physiology & Neurobiology journal homepage: www.elsevier.com/locate/resphysiol

Pleural mesothelium lubrication after hyaluronidase, neuraminidase or pronase treatment Chiara Sironi, Francesca Bodega, Cristina Porta, Emilio Agostoni ∗ Dipartimento di Fisiopatologia e dei Trapianti, Sezione di Fisiologia Umana, Università degli Studi di Milano, Milan, Italy

a r t i c l e

i n f o

Article history: Accepted 6 May 2013 Keywords: Friction coefficient Hyaluronidase Mesothelial glycocalyx Neuraminidase Pronase Triticum vulgaris lectin

a b s t r a c t Coefficient of kinetic friction () of pleural mesothelium has been found to increase markedly after mesothelial blotting and rewetting. This increase disappeared after addition of a solution with hyaluronan or sialomucin, though previous morphological studies showed that only sialomucin occurs in mesothelial glycocalyx. In this research we investigated whether  of rabbit pleural mesothelium increased after hyaluronidase, neuraminidase or pronase treatment. Hyaluronidase and neuraminidase did not increase , though neuraminidase cleaved sialic acid from mesothelial glycocalyx of diaphragm specimens, and removed hystochemical stain of sialic acid from glycocalyx. Sialomucin treated with neuraminidase lowered  of blotted mesothelium, though less than untreated sialomucin; this feature plus lubrication provided by other molecules could explain why  did not increase after neuraminidase. Short pronase treatment (in order to affect only glycocalyx proteins) increased ; this increase was removed by hyaluronan or sialomucin. After pronase treatment  decreased with increase in sliding velocity, indicating a regime of mixed lubrication, as in blotted mesothelium. © 2013 Elsevier B.V. All rights reserved.

1. Introduction Wang (1974, 1985) and Ohtsuka et al. (1997) by transmission electron microscopy found that colloidal iron stain of mesothelial glycocalyx persists after hyaluronidase treatment, but is removed by neuraminidase treatment, which cleaves sialic acid from sialomucin. In line with the earlier proposal of Andrews and Porter (1973), that the polyanionic nature of the surface of the mesothelial glycocalyx may protect from friction, Ohtsuka et al. (1997) concluded that the negative charges of sialomucin produce repulsive forces between facing serosal surfaces, and may, therefore reduce friction. Later on, D’Angelo et al. (2004) measured the coefficient of kinetics friction () between specimens of rabbit visceral and parietal pleura, during oscillatory sliding in vitro at physiological velocities and loads. With Ringer bicarbonate between the mesothelia  was 0.027, and did not change with changes in sliding velocity, consistent with boundary lubrication. Moreover, they found that  increased markedly after having blotted the mesothelial surface with filter paper for 1–2 min, and that this increase was only reduced partially by rewetting the blotted mesothelium with Ringer solution. More recently, with the experimental approach of D’Angelo et al. (2004), we showed that

∗ Corresponding author at: Dipartimento di Fisiopatologia e dei Trapianti, Sezione di Fisiologia Umana, Università degli Studi di Milano, Via Mangiagalli 32, 20133 Milano, Italy. Tel.: +39 02 50315432; fax: +39 02 50315430. E-mail address: [email protected] (E. Agostoni). 1569-9048/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.resp.2013.05.003

addition of sialomucin (25 mg/ml) or hyaluronan (2.5 mg/ml) in Ringer after a standard blotting of the mesothelium brought ␮ back to its pre-blotting value (Bodega et al., 2012). The same effect was obtained by the addition of a mixture of sialomucin (12.5 mg/ml) and hyaluronan (1.25 mg/ml) in Ringer. Moreover, we found that, after washout of the solution with these macromolecules, ␮ with Ringer increased, without reaching its preceding post-blotting value. Finally, transmission electron micrographs of pleural specimens after mesothelial blotting showed that microvilli were partially or largely removed from the mesothelium, consistent with a substantial loss of the macromolecules normally entrapped among them (Bodega et al., 2012). Hence, our findings showed that hyaluronan and sialomucin are able to restore good lubrication in damaged mesothelium, and suggested that sialomucin may be involved in mesothelial lubrication under physiological conditions, while this should not be the case for hyaluronan because of the morphological findings of Wang (1974, 1985), and Ohtsuka et al. (1997). It could, therefore, be interesting to test whether  of pleural mesothelium does not change after hyaluronidase treatment, but increases after neuraminidase treatment. The use of the latter enzyme, however, may be deceptive for various reasons. First, sialic acid (or neuraminic acid), which is placed at the outermost end of the sugar chain of glycoproteins (like sialomucin), is the precursor of several 9-carbon acid sugars in which structural diversities are generated by various substitutions at the 4–9 carbon (Schauer, 1982; Varki and Diaz, 1983; Varki, 1992, 1997). Some of these substitutions can markedly slow or even prevent the release of sialic acid by commonly used neuraminidases (Varki, 1992).

C. Sironi et al. / Respiratory Physiology & Neurobiology 188 (2013) 60–65

The neuraminidase from Streptococcus sanguis has been found to cleave several isomers of sialic acid (Varki and Diaz, 1983), but this enzyme is not commercially available. Moreover, commonly used neuraminidases have a peak activity at pH 5–6, and therefore, their activity at neutral pH may be markedly reduced. Hence, the results obtained by treating the mesothelial specimens with neuraminidase should be taken cautiously. Finally, it could be interesting to test whether  of the pleural mesothelium increases after treatment of the specimens with a protease for a short time, so as to digest only the proteins of the glycocalyx surface without altering the mesothelial cells. The purposes of the present research are, therefore, the following. (1) To test whether  of pleural mesothelium does not change after hyaluronidase treatment of rabbit pleural specimens. (2) To test whether  of pleural mesothelium increases after treatment of the specimens with a commonly used neuraminidase. Moreover, to determine whether sialic acid is cleaved by this enzyme from pleural specimens (like those used to measure ), and from commercial sialomucin (as it should do). These experiments should provide some information that may help to unravel the complex matter mentioned above. (3) To test whether  of pleural mesothelium increases after a short treatment of the specimens with pronase, a broad spectrum protease, that has been used to digest the glycocalyx of mouse (Simionescu et al., 1981, 1985) and frog capillary endothelium (Adamson, 1990).

2. Methods Rib cage, lung, and diaphragm were obtained from 38 rabbits (2.6–3.5 kg b.w.) purchased from “G. Bettinardi”, Momo (Novara). Animal experimentation was authorized by the Ministry of Health by decree N. 60/03A issued according to Order of the Excutive 116/92, in compliance with Directive 86/609/EC. Rabbits were anesthetized with an intravenous injection of 2 ml/kg of a mixture of pentobarbital sodium (Sigma, 12 mg/ml) and urethane (Sigma, 150 mg/ml). They were then heparinised (0.1 mg/kg) and killed by exsanguinations. After removal of the skin and superficial muscles, the antero-lateral sides of the rib cage, the lungs (with closed trachea), and the diaphragm were removed, and kept at room temperature (21–28 ◦ C) in Ringer bicarbonate (in mM: Na+ 139, K+ 5, Ca2+ 1.25, Mg2+ 0.75, Cl− 119, HCO3 − 29, d-glucose 5.6) through which 95% O2 and 5% CO2 was continuously bubbled. The apparatus used to measure the frictional force was that described by D’Angelo et al. (2004). It consists of a sliding platform connected through unextensible threads to the core of a differential transformer, and of a balance arm held stationary at its fulcrum by a force transducer. Tissues specimens to be tested were fixed with the pleural surface facing upwards to the sliding platform, which was driven sinusoidally by an electric motor, and, with the pleural surface facing downwards, to a perspex rod attached to one end of the balance arm. The balance arm was held horizontal, and counterweights added to its other end enabled to change the normal force applied to the tissue from ∼0.5 to ∼8 g, corresponding to a pressure on the mesothelium from ∼0.8 to ∼12.9 cmH2 O. The frictional force on the direction of motion was measured by the force transducer. The coefficient of kinetic friction () was computed as the slope of the relationship between the load and the frictional force. The values of frictional force used were those recorded in the central 40% of the excursion (Bodega et al., 2013). The specimen of rib cage was fixed to the sliding platform by a peripheral frame; alternatively, the specimen of diaphragm was pinned to a flat cork on top of the platform. The specimen of the lung was held over the end of the rod with an O-ring. The sliding velocity was 1.9 cm/s, except otherwise stated.

61

Measurements of  were made at room temperature (21–28 ◦ C) under the following sequence of conditions. (1) Control: Ringer bicarbonate or Krebs phosphate (in mM: Na+ 140.3, K+ 4.4, Ca2+ 2.5, Mg2+ 1.2, Cl− 125.3, SO4 2− 11.2, H2 PO4 − 4.4, d-glucose 6); (2) after enzyme treatment (see below); (3) after washout of the previous solution. In case that  was increased by the enzyme, after its washout  measurements were also made 5 min after addition of a solution with hyaluronan (2.5 mg/ml) or sialomucin (25 mg/ml) on treated mesothelium, and after washout of the solution. In case that  was not significantly increased by the enzyme, in order to test its activity, the enzyme was preincubated with its target molecule, and  measurements were made under this sequence of conditions. (1) Control condition; (2) after mesothelial blotting with filter paper, as previously described (Bodega et al., 2012); (3) after rewetting with Ringer bicarbonare or Krebs phosphate; (4) after addition of enzymatically treated target molecule; (5) after washout of the previous solution; (6) after addition of untreated target molecule; (7) after wash out of the previous solution. 2.1. Enzyme treatment The following enzymes were used: hyaluronidase from Streptomyces hyalurolyticus (Sigma H1136), neuraminidase from Clostridium perfrigens (Sigma N2876), and pronase from Streptomyces griseus (Sigma P8811). In the experiments with hyaluronidase or neuraminidase, after  measurements under control conditions on the lung-diaphragm specimens, the cork with the diaphragm was placed in a 5% CO2 – 95% air humidified incubator (Napco, model 5415) at 37 ◦ C, covered by a thin layer of the enzyme solution, and the piston with the lung specimen was kept on it by mean of an adequate device, and, thus, was also immersed in the enzyme solution. After the incubation time (see below) the piston and the cork were brought back to the apparatus in order to measure . The treatment with pronase was much shorter and, therefore, it was performed at room temperature without removing the piston and the cork from the apparatus. Hyaluronidase was used at a concentration of 30 U/ml in Ringer at pH 7.4 for 90 or 120 min (Knepper et al., 1984). The activity of the enzyme was tested by incubating hyaluronic acid (2.5 mg/ml) with the same concentration of hyaluronidase for 90 or 120 min at 37 ◦ C. This solution was then applied to blotted mesothelium rewetted with Ringer, and its effect on  compared with that of a solution of untreated hyaluronic acid (2.5 mg/ml). Neuraminidase was used at a concentration of 5 U/ml for 60 or 90 min (see below). The activity of neuraminidase from C. perfrigens has a peak at pH 5.5, and decreases to ∼40% at pH 7.0 and to ∼25% at pH 7.4 (Cassidy et al., 1965). For this reason, in preliminary experiments we tested the effect of pH on  of our pleural specimens, and found that after 90 min the lower value of pH at which  did not significantly increase was 7.0. Therefore,  measurements with neuraminidase were performed at pH 7.0, and Krebs phosphate was used because in the incubator it maintains pH at this value for a much longer time than Ringer bicarbonate. On the other hand, it is worth to point out that the same neuraminidase at pH 7.4 (2 U/ml for 30 min at 37 ◦ C) has been found to cleave ∼60% of the sialic acid from tracheal muscles of guinea pig and rat (Kai et al., 1992). Another problem with the neuraminidase used is that, according to the manufacturer, it may have a small protease activity. Hence, we checked whether the neuraminidase used had protease activity. To this end, a 0.5% solution of FITC labelled casein was incubated with 5 or 10 U/ml of neuraminidase for 90 min. To measure the amount of digested protein, casein was precipitated by addition of trichloroacetic acid 0.6 N, and fluorescence intensity in the supernatant was measured with a Perkin-Elmer fluorescence spectrometer MPF4, using an excitation wavelength of 495 nm, and recording emission at 524 nm (Twining, 1984). Pronase was used

62

C. Sironi et al. / Respiratory Physiology & Neurobiology 188 (2013) 60–65

as standard for this assay. With 5 U/ml of neuraminidase no appreciable casein digestion was found, while with 10 U/ml a significant casein digestion occurred. The neuraminidase used cleaves the ˛ 2–3, ˛ 2–6, and ˛ 2–8 linkages between the terminal sialic acid and the glucidic chain (cleavage rate in decreasing order). Neuraminidase from Vibrio cholerae has been successfully used in vivo at physiological pH to remove sialic acid from the luminal surface of rabbit vascular endothelia (Born and Palinski, 1985), though its activity has a peak at pH 5.5, and decreases to about 50% at pH 7.4 (Ada et al., 1961). A neuraminidase with a peak activity at pH 7.2 has been found on the membrane of human erythrocyte (Venerando et al., 1997); however, at present it is not available. The activity of the neuraminidase used was tested by incubating sialomucin (25 mg/ml) with 5 U/ml of neuraminidase for 60 or 90 min at 37 ◦ C. In order to separate the unbound sialic acid, the solution was then filtered through ultrafiltration low-binding cellulose ultrafiltration membranes (100 kDa nominal molecular weight limit, Millipore) by centrifuging at 5000 × g for 10 min at 4 ◦ C. The retentate of the above filtration (diluted with a volume of Krebs at pH 7 equal to the filtered volume) was applied to blotted mesothelium rewetted with Krebs at pH 7.4, and its effect on  compared with that of a solution of untreated sialomucin. Sialic acid concentration in the above ultrafiltrate was fluorimetrically measured. To this end 0.2 ml of the solution was added to an equal volume of 5 ␮M 3,5-di-aminobenzoic acid (DABA, Sigma 32770) in 0.125 N HCl, and the mixture was heated at 110 ◦ C for 16 h. A 50 ␮l sample was then transferred into 1.5 ml of 0.05 N HCl and its fluorescence was determined with a Perkin-Elmer fluorescence spectrometer MPF4, using an excitation wavelength of 405 nm and recording emission at 530 nm (Hess and Rolde, 1964; Born and Palinski, 1985). The amount of sialic acid was also measured in the ultrafiltrate of a solution incubated without neuraminidase in order to measure the small, but not negligible, amount of free sialic acid occurring in commercial sialomucin. The amount of sialic acid cleaved by the enzyme was then given by that measured in the ultrafiltrate of the solution with neuraminidase (total) minus that in the ultrafiltrate of the solution without neuraminidase (free). Because commercial sialomucin (obtained from bovine submaxillary gland) contains various isomers of sialic acid (Schauer, 1982), the activity of neuraminidase was also tested on diaphragm specimens (like those used for  measurements) incubated for 90 min at 37 ◦ C in Krebs phosphate at pH 7.0 without (control) or with 5 U/ml of neuraminidase. To this end the diaphragm specimen was placed on a perspex disc with the pleural mesothelium facing up, and fastened by a peripheal perspex ring that was screwed on the underlying disc. The ring was 0.6 cm thick, and its internal diameter was 1 cm, thus leaving 0.78 cm2 of mesothelial surface uncovered; 0.1 ml of the incubating solution was poured on this mesothelial surface. The sialic acid concentration in the incubating solutions was then fluorimetrically measured to obtain the amount of cleaved sialic acid (total – free, see above). The occurrence of sialic acid on the surface of pleural glycocalyx and its disappearance after neuraminidase treatment was also ascertained by histochemistry. To this end lectin from Triticum vulgaris was used, which is specific for sialic acid and for Nacetylglucosamine (Hjelle et al., 1991). Specimens of pleura and underlyng lung were incubated for 90 min with Krebs phosphate at pH 7.0 or with 5 U/ml of neuraminidase in Krebs phosphate at pH 7.0. They were then fixed in 4% paraformaldehyde in 120 mM phosphate buffer (in mM: Na+ 140, H2 PO4 − 100, HPO4 2− 20) for 20 min and then left in PBS (in mM: Na+ 155, HPO4 2− 7.7, H2 PO4 − 2.7, Cl− 137) overnight. The fixed blocks of lung were washed in PBS, dehydrated by passing through graded alcohols, and then embedded in paraffin (overnight at 60 ◦ C). Seven ␮m thick sections were cut by mean of a microtome (RM 2125RT, Leica, Milan, Italy),

collected on poly-l-lysine slides (Poly-prep slides, Sigma), dried at 37 ◦ C overnight, and kept at room temperature until further processing. Before staining, sections were placed at 60 ◦ C for 30 min, dewaxed in xylene, and rehydrated by passing through graded alcohols. Section were washed three times in PBS and then incubated at room temperature for 1 h with fluorescein-labelled lectin from Triticum vulgaris (Sigma L4895), diluted 1:1000 in PBS. After incubation the unbound lectin was removed by washing three times with PBS. Sections were then examined for fluorescence with a Zeiss Axioscope 40 microscope. Pronase was used at a concentration of 0.1 mg/ml (0.58 U/ml) for 5 min. It has been shown on frog mesenteric capillaries that at this concentration and for short incubation time (1 min) pronase digestion is associated with partial digestion of the endothelial glycocalyx, but is not accompanied by changes in the dimensions of the intercellular cleft (Adamson, 1990). Moreover, even at a greater concentration (4–6 U/ml) and for a longer period (10–20 min), the use of pronase to digest the endothelial glycocalyx, does not seem to damage the endothelial cells of the capillaries of mouse pancreas and intestinal mucosa (Simionescu et al., 1981), and of the capillaries of mouse hearth and diaphragm (Simionescu et al., 1985). Anyway we checked whether the pattern of distribution of cellular nuclei (and, hence, of cells) was not disrupted by short treatment with this enzyme. To this end, specimens of tendinous diaphragm untreated or treated with pronase for 5 min were fixed with 4% paraformaldehyde for 10 min. They were then washed with PBS, and incubated for 5 min with DAPI (0.057 ␮g/ml, Sigma) to mark cellular nuclei. Samples were then placed on microscope slides with some drops of 50% glycerol in PBS and examined for fluorescence with a Zeiss Axioscope 40 microscope. Linear regressions between frictional force and load were computed with the least squares method and statistical assessment was made by covariance analysis. The results are presented as mean ± SE. Statistical significance of group mean values was assessed by analysis of variance. The level of significance was taken at P < 0.05.

3. Results and discussion 3.1. Hyaluronidase The coefficient of kinetic friction () of specimens of rabbit pleura after the addition of 30 U/ml of hyaluronidase in Ringer bicarbonate solution for 90 or 120 min at 37 ◦ C was 0.032 ± 0.003, and 0.034 ± 0.003 after washout of the enzyme with Ringer: both values were not significantly greater (P > 0.05) than that with initial Ringer in these specimens (0.031 ± 0.003; N = 10). No difference occurred between the specimens in which the enzyme was applied for 90 min and those in which it was applied for 120 min. The activity of the enzyme was tested by adding to postblotting Ringer specimens hyaluronic acid (2.5 mg/ml) that had been incubated with the same concentration of hyaluronidase for 90 or 120 min at 37 ◦ C. The value of  after the addition of treated hyaluronic acid was 0.064 ± 0.008, not significantly lower (P > 0.05) than that of post-blotting Ringer in the same specimens (0.068 ± 0.006, N = 10). Hence, the enzyme was active, but was unable to affect the lubrication of pleural mesothelium under the initial control condition. If hyaluronic acid (2.5 mg/ml) was added to the same post-blotting Ringer specimens treated with digested hyaluronic acid, and washed out with Ringer,  decreased (P < 0.01) to 0.032 ± 0.004, i.e. it became similar to that of the initial control condition (0.028 ± 0.002). The finding that  of the pleural mesothelium does not increase after treatment with hyaluronidase is in line with the morphological findings of Wang (1974, 1985) and Ohtsuka et al. (1997) showing

C. Sironi et al. / Respiratory Physiology & Neurobiology 188 (2013) 60–65

63

Table 1 Effect of neuraminidase treated or untreated sialomicin (25 mg/ml) on coefficient of kinetic friction () of pleural mesothelium in Krebs solution after blotting with filter paper and rewetting with Krebs. Condition

1 Krebs pH 7.4

2 Blotting

3 Krebs pH 7.4

4 Treated sialomucin in Krebs pH 7.0

5 Krebs pH 7.4

6 Sialomucin in Krebs pH 7.4

7 Krebs pH 7.4

 ±SE

0.027 0.002

0.202 0.017

0.089 0.015

0.053 0.006

0.078 0.011

0.038 0.005

0.069 0.010

 ±SE t P

 3–4

 4–1

 5–4

 5–6

 4–6

 6–1

 7–6

 3–7

0.036 0.010

0.026 0.005

0.025 0.006

0.040 0.009

0.015 0.003

0.011 0.004

0.031 0.008

0.020 0.006

3.49 <0.01

5.12 <0.01

4.05 <0.01

4.58 <0.01

4.46 <0.01

2.84 <0.05

3.94 <0.01

3.12 <0.02

Values are means ± S.E.; N = 10 (6 lung – muscular diaphragm, 3 lung – intercostal, and 1 lung – tendinous diaphragm).

that hyaluronic acid does not occur on the surface of the glycocalyx of the mesothelium. It has been shown that mesothelial cells in vitro are surrounded by a coat of hyaluronic acid produced by these cells (Heldin and Pertoft, 1993); this, however, does not necessarily imply that hyaluronic acid is on the surface of the mesothelial glycocalyx. Hyaluronic acid occurs in the endothelial glycocalyx (which is ∼0.5 ␮ thick), and can be removed in vivo by hyaluronidase (Henry and Duling, 1999). Though hyaluronic acid does not seem to be located on the surface of the mesothelial glycocalyx, it is intriguing that its addition after pleural mesothelium blotting (Bodega et al., 2012) brings  back to control condition, and restores boundary lubrication. The concentration of hyaluronic acid in rabbit pericardial liquid has been found to be 82 ␮g/ml (Honda et al., 1986), and that in pleural liquid ∼1 ␮g/ml (Wang and Lai-Fook, 1998). Both concentrations are much smaller than that required to bring  back to its control value after mesothelial blotting. Hence, despite hyaluronic acid acts as a good boundary lubricant when applied to damaged mesothelium, it is as if in vivo its concentration is kept low for other reasons. 3.2. Neuraminidase The value of  of pleural mesothelium after addition of 5 U/ml of neuraminidase in Krebs phosphate at pH 7 for 60 or 90 min at 37 ◦ C was 0.025 ± 0.002, and 0.028 ± 0.001 after washout with Krebs at pH 7.4: both values are not significantly greater (P > 0.05) than that of initial Krebs at pH 7 in these specimens (0.025 ± 0.002, N = 10). No difference occurred between the specimens in which the enzyme was applied for 60 min and those in which it was applied for 90 min. The activity of the enzyme in Krebs at pH 7 was tested by fluorimetric measurements of sialic acid in filtered samples of a solution in which sialomucin (25 mg/ml) had been incubated without or with 5 U/ml of neuraminidase for 90 min at 37 ◦ C (see Methods). The amount of sialic acid cleaved by the enzyme (i.e. total – free) from 3.75 mg of sialomucin was 0.65 mg, i.e. ∼17% of the sialomucin. This corresponds to all the sialic acid bound in the commercial sialomucin (which is obtained from bovine submaxillary gland). Hence, the neuraminidase used was fully active at least on this sialomucin. The retentate of the above filtration (i.e. the glycoprotein without its sialic acid) was diluted with a volume of Krebs at pH 7 equal to the filtered volume; this solution was added to post-blotting specimens in Krebs at pH 7.4. The value of  after the addition of this solution was 0.053 ± 0.006, i.e. smaller (P < 0.01) than that in post-blotting Krebs (0.089 ± 0.015), but greater (P < 0.01) than that in pre-blotting Krebs (0.027 ± 0.002, N = 10, see Table 1). After washout of treated sialomucin solution with Krebs at pH 7.4,  increased (P < 0.01) to 0.078 ± 0.011, and after addition of untreated sialomucin (25 mg/ml) it decreased (P < 0.01) to 0.038 ± 0.005, which is lower (P < 0.01) than that with treated sialomucin solution, and nearly back to its initial pre-blotting value, though still

a little higher (P < 0.05, Table 1). After washout of the sialomucin with Krebs at pH 7.4,  increased (P < 0.01) to 0.069 ± 0.010, but was lower (P < 0.02) than in post-blotting Krebs at pH 7.4 (Table 1), as previously shown with Ringer (Bodega et al., 2012). Hence, this glycoprotein without its sialic acid maintains a lubricating effect, though smaller than that of untreated sialomucin. Fluorimetric measurements of sialic acid were also performed in samples of Krebs phosphate solution at pH 7 without or with 5 U/ml of neuraminidase incubated for 90 min at 37 ◦ C on the pleural surface of 6 diaphragmatic specimens (see Methods). The amount of sialic acid cleaved by the enzyme (i.e. total – free) was 0.0515 ± 0.0049 mg. Hence, the neuraminidase used is active to some extent also on the sialomucin of the glycocalyx of the pleural mesothelium of diaphragm specimens (like those used for  measurements). The above amount corresponds to 0.066 mg/cm2 of the macroscopic mesothelial surface of the diaphragmatic specimen (0.78 cm2 , see Methods). On the other hand, because of the microvilli, the corresponding microscopic surface of the glycocalyx may be thought to be much larger than the macroscopic surface of the pleural mesothelium. This difference, however, should not be large because the macromolecules of the glycocalyx are entrapped among the microvilli (Andrews and Porter, 1973); moreover, according to an electron microscope image of (Michailova, 2004, Fig. 3F) a strip-like structure covers the granulo-filamentous material and microvilli of mesothelial cells. If this were the case the difference between microscopic and macroscopic surface should be relatively small being essentially due to the bumpiness of the surface related to cell shape, and to the tension applied to the specimen. This small difference may, therefore, occur also in tissues without microvilli. The amount of sialic acid cleaved by neuraminidase from the glycocalyx of rabbit vascular endothelium per unit of macroscopic surface has been found to be 0.002 mg/cm2 (Born and Palinski, 1985), i.e. much less than that cleaved from the glycocalyx of rabbit pleural mesothelium. Only part of this difference may be due to the lower enzymatic concentration used in the vascular endothelium, the rest suggests that the amount of sialic acid (and, hence, of sialomucin) per unit of glycocalyx surface is greater in pleural mesothelium than in vascular endothelium. Fig. 1 shows fluorescent microscopy images of sections of pleura and underlying lung parenchima obtained from specimens preincubated for 90 min at 37 ◦ C in Krebs phosphate at pH 7.0 without (A) or with (B) 5 U/ml of neuraminidase. Sialic acid was stained with fluorescein-labelled lectin from Triticum vulgaris (green). Lectin labelling occurs only on control mesothelial surface: hence neuraminidase has removed most of the sialic acid bound to the sialomucin on the surface of the glycocalyx of pleural mesothelium. Our hystochemical finding fits with the morphological findings of Wang (1974, 1985), and Ohtsuka et al. (1997) showing that sialomucin occurs on the surface of the mesothelial glycocalyx. On the other hand, it seems to contrast with the lack of

64

C. Sironi et al. / Respiratory Physiology & Neurobiology 188 (2013) 60–65

Coefficient of kinetic friction

0.14 0.12 0.10 0.08 0.06 0.04 0.02 0.00

0

1

2

3

4

5

Velocity, cm/s

Fig. 1. Fluorescent microscopy images of sections of pleura and underlying lung obtained from specimens preincubated without (A) or with (B) neuraminidase: sialic acid was stained with fluorescein-labelled lectin from Triticum vulgaris (green). Lectin labelling occurs only on control mesothelial surface. Alveolar and pleural tissues are faintly outlined by autoflorescence. Bar = 20 ␮m.

Fig. 2. Coefficient of kinetic friction () versus sliding velocity during reciprocating movements of visceral against parietal pleura after 5 min pronase treatment. N = 9 for each circle; verticals bars indicate SE. Dotted line shows for comparison the relationship previously obtained after mesothelial blotting and rewetting with Ringer (Bodega et al., 2013).

3.3. Pronase increase in  of the pleural mesothelium after neuraminidase treatment, despite the fact that this enzyme cleaves all sialic acid of commercial sialomucin, and likely most of that of the sialomucin occurring in the glycocalyx of pleural mesothelium of diaphragm specimens (see above). Like hyaluronic acid, sialomucin applied to damaged mesothelium has been shown to bring  back to its control value when (Bodega et al., 2012). Since hyaluronic acid does not occur on the surface of mesothelial glycocalyx, while sialomucin does, it seemed like that the good lubrication occurring in the pleural mesothelium was mainly due to sialomucin of the glycocalyx. However, the lack of increase in  of pleural mesothelium after neuraminidase treatment suggests that other molecules of the glycocalyx contribute to the good lubrication of the pleural mesothelium. Hence, this matter cannot be solved until more complete information on the components of the glycocalyx of pleural mesothelium is available. At present, it is noteworthy that commercial sialomucin treated with neuraminidase (i.e. without all its sialic acid) maintains a substantial lubricating effect in postblotting mesothelium rewetted with Krebs at pH 7.4 (Table 1, Columns 3–4). This feature plus the contribution of other lubricating molecules could tentatively explain why  of pleural mesothelium does not increase after neuraminidase treatment. There is another point that deserves a comment. After washout of treated sialomucin with Krebs at pH 7.4 and the addition of untreated sialomucin (Table 1, Columns 5–6), the decrease in  is greater than that produced by treated sialomucin, but insufficient to reach a value not significantly greater than the control one (Table 1,  6 − 1). Instead, the addition of untreated sialomucin to post-blotting Ringer was able to bring  back to its control value (Bodega et al., 2012). To try to understand the cause of this difference we made a series of  measurements on 9 couples of specimens (5 lung – intercostal, 3 lung – muscular diaphragm and 1 lung – tendinous diaphragm) in which untreated sialomucin (25 mg/ml) in Krebs at pH 7.4 was applied to postblotting specimens rewetted with Krebs at pH 7.4. The value of  decreased from 0.086 ± 0.005 to 0.042 ± 0.003, which is close to, but still significantly higher (P < 0.02) than the control value (0.029 ± 0.002). Hence, the finding that the addition of untreated sialomucin in post-blotting Krebs is unable to bring  back to its control value seems related to the use of Krebs instead of Ringer.

After the addition to the specimens of 0.1 mg/ml of pronase in Ringer bicarbonate for 5 min at room temperature  increased (P < 0.01) from 0.031 ± 0.002 (N = 14) to 0.063 ± 0.007, and 0.071 ± 0.009 after enzyme washout with Ringer. In 6 of these specimens the addition of hyaluronic acid (2.5 mg/ml) after pronase treatment decreased (P < 0.01)  to 0.024 ± 0.004, which is not significantly different (P > 0.05) from that in initial Ringer (0.029 ± 0.004). In the other 8 specimens the addition of sialomucin (25 mg/ml) after pronase treatment decreased (P < 0.01)  to 0.044 ± 0.008, which is not significantly greater (P > 0.05) than that in initial Ringer (0.032 ± 0.002). Because the increase in  produced by the addition of pronase was nearly as marked as that occurring in post-blotting Ringer (Bodega et al., 2012), in 9 of the above mentioned specimens we also determined  in post-pronase Ringer at a lower and a higher sliding velocity (0.9 and 4.7 cm/s, respectively) to check whether it decreased with the increase in velocity, as it does in post-blotting Ringer (Bodega et al., 2013). The value of  decreased from 0.100 ± 0.018 at 0.9 cm/s to 0.070 ± 0.013 at 1.9 cm/s (P < 0.05), and to 0.049 ± 0.010 at 4.7 cm/s (P < 0.01). Hence, after the addition of pronase to the specimens,  decreases with the increase in sliding velocity (Fig. 2). Despite the marked decrease in  that occurred with the increase in sliding velocity,  in post-pronase Ringer at the highest velocity was still greater (P < 0.05) than that in initial Ringer (0.029 ± 0.003), which is independent of sliding velocity (D’Angelo et al., 2004; Bodega et al., 2013). Given the short period of treatment with pronase, its activity should be limited to the glycocalyx without affecting the mesothelial cells. Indeed, the distribution pattern of the nuclei of the mesothelial cells after pronase treatment was not disrupted. Hence, the marked increase in  produced by the addition of pronase should be mainly due to the digestion of the proteinic component of the glycoproteins of the glycocalyx. The changes in mesothelial surface produced by this treatment should be markedly different from that produced by mesothelial blotting because the latter involves partial or large removal of the microvilli and of the glycocalyx (Bodega et al., 2012). Nevertheless, even after short pronase treatment the marked increase in  decreased significantly with the increase in sliding velocity (Fig. 2), indicating that also in this case there is a regimen of mixed lubrication (boundary

C. Sironi et al. / Respiratory Physiology & Neurobiology 188 (2013) 60–65

and elasto-hydrodynamic). In post-blotting Ringer the mesothelial surface is much more rough than under control conditions, and in several points the amount of liquid between the opposed surfaces should be greater than before blotting. Its seems likely, therefore, that when the mesothelial surfaces are made to slide, the shear induced hydrodynamic pressure generates some areas of hydrodynamic lubrication, which increase with the increase in sliding velocity, while those of boundary lubrication decrease (Bodega et al., 2013). In post-pronase Ringer the roughness of the mesothelial surface should not be increased so much, but the changes undergone by the lubricating macromolecules likely decrease the areas of boundary lubrication, and favour the onset of elasto-hydrodynamic areas with the increase in sliding velocity. Indeed, the addition of sialomucin or hyaluronic acid after pronase treatment brought back  to control value (see above), and, hence, to a condition of boundary lubrication which is independent from sliding velocity (D’Angelo et al., 2004; Bodega et al., 2013). Acknowledgments We thank Prof. G. Tettamanti for helpfull discussion on neuraminidases, Prof. E D’Angelo and Dr. M. Pecchiari for the apparatus used to measure the coefficient of kinetic friction. Moreover, we thank R. Galli for his skilful technical assistance. References Ada, G.L., French, E.L., Lind, P.E., 1961. Purification and properties of neuraminidase from Vibrio cholerae. Journal of General Microbiology 24, 409–425. Adamson, R.H., 1990. Permeability of frog mesenteric capillaries after partial pronase digestion of the endothelial glycocalyx. Journal of Physiology 428, 1–13. Andrews, P.M., Porter, K.R., 1973. The ultrastructural morphology and possible functional significance of mesothelial microvilli. Anatomical Record 177, 409–426. Bodega, F., Pecchiari, M., Sironi, C., Porta, C., Arnaboldi, F., Barajon, I., Agostoni, E., 2012. Lubricating effect of sialomucin and hyaluronan on pleural mesothelium. Respiratory Physiology and Neurobiology 180, 34–39. Bodega, F., Sironi, C., Porta, C., Pecchiari, M., Zocchi, L., Agostoni, E., 2013. Mixed lubrication after rewetting blotted mesothelium. Respiratory Physiology and Neurobiology 185, 369–373. Born, G.V., Palinski, W., 1985. Unusually high concentrations of sialic acids on the surface of vascular endothelia. British Journal of Experimental Pathology 66, 543–549. Cassidy, J.T., Jourdian, G.W., Roseman, S., 1965. The sialic acids. VI. Purification and properties of sialidase from Clostridium perfringens. Journal of Biological Chemistry 240, 3501–3506.

65

D’Angelo, E., Loring, S.H., Gioia, M.E., Pecchiari, M., Moscheni, C., 2004. Friction and lubrication of pleural tissues. Respiratory Physiology and Neurobiology 142, 55–68. Heldin, P., Pertoft, H., 1993. Synthesis and assembly of the hyaluronan-containing coats around normal human mesothelial cells. Experimental Cell Research 208, 422–429. Henry, C.B., Duling, B.R., 1999. Permeation of the luminal capillary glycocalyx is determined by hyaluronan. American Journal of Physiology: Heart and Circulatory Physiology 277, H508–H514. Hess, H., Rolde, E., 1964. Fluorimetric assay of sialic acid in brain gangliosides. Journal of Biological Chemistry 239, 3215–3220. Hjelle, J.T., Golinska, B.T., Waters, D.C., Steidley, K.R., Miller, M.A., McCarroll, D.R., Dobbie, J.W., 1991. Lectin staining of peritoneal mesothelial cells in vitro. Peritoneal Dialysis International 11, 307–316. Honda, A., Ohashi, Y., Mori, Y., 1986. Hyaluronic acid in rabbit pericardial fluid and its production by pericardium. FEBS Letters 203, 273–278. Kai, H., Makise, K., Matsumoto, S., Ishii, T., Takahama, K., Isohama, Y., Miyata, T., 1992. The influence of neuraminidase treatment on tracheal smooth muscle contraction. European Journal of Pharmacology 220, 181–185. Knepper, P.A., Farbman, A.I., Telser, A.G., 1984. Exogenous hyaluronidases and degradation of hyaluronic acid in the rabbit eye. Investigative Ophthalmology & Visual Science 25, 286–293. Michailova, K.N., 2004. Mesothelial lamellar bodies in norm and experimental conditions. Transmission and scanning electron microscopic observations on the peritoneum, pleura and pericardium. Anatomy and Embryology 208, 301–309. Ohtsuka, A., Yamana, S., Murakami, T., 1997. Localization of membrane-associated sialomucin on the free surface of mesothelial cells of the pleura pericardium and peritoneum. Histochemistry and Cell Biology 107, 441–447. Schauer, R., 1982. Chemistry, metabolism, and biological functions of sialic acids. Advances in Carbohydrate Chemistry and Biochemistry 40, 131–234. Simionescu, M., Simionescu, N., Santoro, F., Palade, G.E., 1985. Differentiated microdomains of the luminal plasmalemma of murine muscle capillaries: segmental variations in young and old animals. The Journal of Cell Biology 100, 1396–1407. Simionescu, M., Simionescu, N., Silbert, J.E., Palade, G.E., 1981. Differentiated microdomains on the luminal surface of the capillary endothelium. II. Partial characterization of their anionic sites. The Journal of Cell Biology 90, 614–621. Twining, S.S., 1984. Fluorescein isothiocyanate-labeled casein assay for proteolytic enzymes. Analytical Biochemistry 143, 30–34. Varki, A., 1992. Diversity in the sialic acids. Glycobiology 2, 25–40. Varki, A., 1997. Sialic acids as ligands in recognition phenomena. FASEB Journal 11, 248–255. Varki, A., Diaz, S., 1983. A neuraminidase from Streptococcus sanguis that can release O-acetylated sialic acids. Journal of Biological Chemistry 258, 12465–12471. Venerando, B., Fiorilli, A., Croci, G.L., Tettamanti, G., 1997. Presence in human erythrocyte membranes of a novel form of sialidase acting optimally at neutral pH. Blood 90, 2047–2056. Wang, N.S., 1974. The regional difference of pleural mesothelial cells in rabbits. American Review of Respiratory Disease 110, 623–633. Wang, N.S., 1985. Mesothelial cells in situ. In: Chrétien, J., Bignon, J., Hirsh, A. (Eds.), The Pleura in Health and Disease. Dekker, New York, pp. 23–42. Wang, P.M., Lai-Fook, S.J., 1998. Effects of ventilation on hyaluronan and protein concentration in pleural liquid of anesthetized and conscious rabbits. Lung 176, 309–324.